# Changelog
Source: https://docs.babbily.com/changelog/overview
What's new in Babbily.
Updates below describe availability at the time of each release. For today's options, use your model picker and check [current model transitions](/features/model-capabilities#sunsetting-and-removed-models).
## More relevant pictures from web search
Web search now adds pictures only when they fit the question. For the first question in a chat, Babbily can include up to three pictures of a subject you can see, such as a place, product, or landmark. In follow-up questions, Babbily adds search pictures only when you ask to see them, for example `Show me photos of the Eiffel Tower at night`. [Learn how to ask for pictures](/features/visual-replies-in-chat#ask-for-pictures-from-web-search).
Search pictures now appear after the written answer starts instead of before it, and Babbily leaves out logos, icons, ads, tiny thumbnails, and pictures that do not match your question. Deep Research reports do not include search pictures.
## GPT-6 Astra, updated image generation, and more control over ongoing work
**GPT-6 Astra** is available for complex reasoning, coding, and extended research, with **Fast mode** and reasoning effort through **Max**. [Compare models and controls](/features/model-capabilities).
**GPT Image 2.5 Flare** is available for image generation and editing. The chat **Image generation** tool also uses it for new images from text prompts. **GPT Image 2** remains selectable as **Sunsetting**. [Choose the image model or chat tool](/features/image-and-video-generation#choose-the-image-model-or-chat-tool).
**Organize existing threads.** Use **Add to project**, **Move to project**, or **Remove from project** on your own eligible threads within their current workspace. You can also create a project from the thread actions. [Learn how project context changes](/features/projects#add-an-existing-thread).
**Review what Babbily remembers.** Older items may appear under **Needs review**, with individual and bulk approval or forget controls. Memory save feedback makes outcomes clearer. Forgetting memory leaves the original chat history intact. [Review memory controls](/features/memory).
**Find earlier discussions and product help.** In eligible chats, **Auto** mode can look up earlier conversations when memory is enabled. Babbily can also look up help-center instructions with **Babbily Docs**. [Check lookup availability](/features/memory#refer-to-past-chats).
Long-chat context handling has also been updated. Summaries help keep work moving, but can omit details; restate exact requirements when needed. [Understand the context window](/features/context-window).
## Fast mode for GLM 5.3 Flash
Select **GLM 5.3 Flash**, then enable **Fast mode** to use GLM 5.3 Fast. The model name stays the same in the picker, and reasoning remains an on/off control. File support can differ in Fast mode; check the composer before attaching files. [Review Fast mode](/features/model-capabilities#use-fast-mode).
## Edit earlier messages and find saved work more easily
Use **Edit message** to revise one of your earlier prompts and generate a new reply. Sending the edit removes all later messages, so save any work you still need first. [Review editing and attachment limits](/features/chat-and-models#edit-an-earlier-message).
The Library has updated browsing, sorting, and item actions across desktop and mobile, including **Files & Artifacts**. [Find saved work](/features/library-and-sharing).
Notification setup now includes Home Screen guidance for iPhone and iPad. Passkey setup prompts also better reflect whether you already have a registered passkey. [Manage notifications](/features/account-and-settings#notifications-on-iphone-and-ipad) or [manage passkeys](/features/account-access#use-passkeys).
## Gemini 3.8 Flash
**Gemini 3.8 Flash** is available for chat and **Translate**. It replaces Gemini 3.7 Flash in Translate; Gemini 3.7 Flash remains selectable as **Sunsetting** in chat. [Review current translation models](/features/translate#available-translation-models).
## New models, Projects, Scheduled Tasks, and the new Plugins experience
Babbily adds three models, refreshes older model options, gives ongoing work a dedicated home, runs recurring prompts for paid users, and connects apps through a new Plugins experience.
## New models and availability
**Claude Fable 5.1** is available for long-running coding, knowledge work, research, and precise instruction following. **GLM 5.3 Flash** supports visual coding, tool use, and work across code, documents, and images. **Gemini 3.7 Flash** supports efficient coding, long-context reasoning, and multimodal work, and is also available in **Translate**. All three have a 1-million-token context window.
**Four predecessors move to Sunsetting.** Claude Fable 5, GLM 5.2, Gemini 3.6 Flash, and Gemini 3.5 Flash Lite remain selectable in chat during the transition. Choose their newer alternatives for new work. GLM 5.2 retains **Fast mode**; the three new models do not have a separate Fast mode toggle.
**Older sunsetting options leave the picker.** Fifteen previously sunsetting models, including Claude Opus 4.8, Grok 4.5, Muse Spark 1.1, Seedance 2.0, and FLUX 2 Pro, are no longer selectable in the picker. See the [complete list and current alternatives](/features/model-capabilities#sunsetting-and-removed-models).
**Translate moves to Gemini 3.7 Flash.** Gemini 3.6 Flash and Gemini 3.5 Flash Lite are no longer options in Translate. GPT-5.6 Sol and Claude Haiku 4.5 remain available there.
Review new models, reasoning effort, Fast mode, and current alternatives.
Check the current Translate options and Gemini transition.
## What's new
**Keep related work together in Projects.** Create a Project in your **Personal** or team workspace to keep its chats, instructions, knowledge files, and project memory together. Project instructions shape every new chat, and reusable knowledge can include supported PDF, DOCX, text, Markdown, CSV, and JSON files. Team members can collaborate on shared project content while each chat remains continuable only by its author.
**Run recurring prompts with Scheduled Tasks.** Paid plans can use **Scheduled** for manual, hourly, daily, weekday, weekly, or monthly prompts. Choose a model, time zone, and any connected Plugins the task may read from. Each run uses the API usage budget of its workspace and leaves its answer in a new chat under **Recent runs**.
**Connect apps through Plugins.** **Connectors** is now **Plugins** throughout Babbily. The supported Plugin catalog now uses Composio for authorization, while Babbily continues to ask for approval before attended write actions run. Availability can vary by your app permissions, account state, and service status.
## What you need to do
Former Smithery connections were retired rather than transferred. Open **Settings > Plugins**, choose each app you still want to use, and complete the new Composio authorization flow. Your previous connections can appear under **Previous connection history**, but their credentials were not moved.
If an older Scheduled Task shows **Plugin reselection required**, edit the task after reconnecting its apps. Choose the current Plugins it may use—even if the correct choice is none—and save before you run or resume it.
Keep chats, instructions, reference files, and project memory together.
Automate recurring prompts and review each run.
Review the current app catalog, authorization, and approvals.
Fix pop-ups, authorization, reconnection, and reselection issues.
## We retired Saved Prompts
Saved Prompts is no longer available in Babbily. We made this change after seeing low usage and as Babbily continues to evolve toward more agentic ways of working.
## What this means
**Saved Prompts is no longer available.** You can no longer create, edit, or open Saved Prompts in Babbily.
**Use memory for context you want reused.** Memory can carry stable preferences, work context, recurring projects, tone, audience, and formatting choices into future chats. This reduces how much setup you need to repeat.
**Use prompt patterns for repeatable tasks.** Start with a consistent goal, context, audience, constraints, and output format. Keep task-specific instructions in the chat so you can adjust them as the work changes. You can also use a Skill when one fits your workflow.
Save stable context you want Babbily to use in future chats.
Structure reusable instructions for common tasks.
Apply reusable instructions and workflows when a Skill fits your task.
## More reliable Finance research and a new Deep Research Agent
Babbily now keeps Finance research grounded across longer conversations and can build broad, source-backed reports with a new Deep Research Agent.
## What's new
**Keep Finance context through follow-ups.** Start from a market overview, company, earnings event, or prediction market and continue asking questions without reselecting Finance research. Supported Finance threads preserve the page context and your tool choice across follow-ups, page reloads, and **Retry**.
**Control Finance research with Auto, Manual, or Off.** In a supported Finance thread, **Auto** uses Finance research automatically. **Manual** starts with **Finance research** enabled for that thread, and you can remove it. **Off** keeps the reply model-only.
**Get stronger source coverage for financial questions.** Finance research now handles live quotes, historical periods, earnings, peer comparisons, and multi-company filing research more consistently. It checks requested coverage, keeps citations with the answer they support, and uses primary filings when available.
**See clear limits instead of unsupported figures.** When a company does not report a requested metric separately, Babbily explains what is unavailable and can suggest a related figure that is reported. Ambiguous company names or missing evidence may produce a clarification or limitation instead of a guess.
**Build full reports with the Deep Research Agent.** Choose **Deep research** for a broad question. Babbily plans sub-questions, opens the outline in Canvas, gathers and verifies sources, and streams a cited report while it works. Chat keeps a short summary, and the full report stays in Canvas for review, editing, and export.
**Keep useful work when a long run is interrupted.** If research reaches its time limit, Babbily keeps the report and tells you its coverage is partial. If report writing reaches its limit, the report is preserved and labeled incomplete. If a selected plugin action needs approval, the request appears inline in the conversation; the report pauses and can continue after you approve it.
Ask source-aware questions from Finance and understand how each tool mode behaves.
Plan a multi-source question, follow progress in Canvas, and review the cited report.
Edit, version, export, or continue from a completed or partial research report.
Compare Auto, Manual, and Off for research and other built-in tools.
## Claude Opus 5 is now available with Fast mode
Claude Opus 5 is now available in the Babbily model picker. Use it for agentic coding, professional knowledge work, long-horizon reasoning, and vision-enabled work across supported files and images.
## What's new
**Work with a 1-million-token context window.** Claude Opus 5 can keep more conversation, file, and tool context in view for extensive codebases, long documents, and sustained multi-step work.
**Turn on Fast mode for quicker replies.** Select Claude Opus 5, then choose **Fast mode** using the lightning control near the composer. Fast mode uses a lower-latency response path while Claude Opus 5 remains your selected model.
**Plan for higher usage in Fast mode.** Fast mode can respond faster, but it uses more of your API usage budget. Turn it off when you want the standard Claude Opus 5 response path.
**Move new Opus work forward.** Claude Opus 4.8 remains available with a **Sunsetting** label during the transition. Choose Claude Opus 5 for new Opus chats.
Decide when Claude Opus 5, Fast mode, or another model fits your task.
Find the Fast mode control and understand its API usage budget impact.
## Workspaces, team collaboration, and two new models
Babbily now keeps personal and team work in distinct workspaces. You can switch between **Personal** and every team workspace you belong to, collaborate with teammates without making work public, and keep an individual plan alongside a team plan.
## What's new
**Switch between workspaces.** Open your account menu to choose **Personal** or a team workspace. The workspace pill in the sidebar shows where you are working. Chats, Library items, and memory stay with the workspace where you created them, so switching workspaces does not move existing work.
**Share work with your team.** From a team workspace, you can share a chat or supported saved media with specific teammates or everyone in the team. Recipients find it under **Shared with you** in Library. Shared team chats are read-only, and access lasts only while the recipient remains a team member.
**Use Shared team memory.** A team owner can turn **Shared team memory** on from **Settings > Team**. When it is on, useful facts from team-workspace chats can become shared context that current members can use in team chats. Personal memory remains separate. Authors and the team owner can delete eligible shared entries, and turning the setting off stops new additions.
**Keep personal and team plans side by side.** Joining or creating a team no longer requires your individual plan to disappear. Your Personal workspace and team workspace show their own plan context, and owners manage a team plan from the team workspace they own. If a team plan becomes inactive, Babbily labels that workspace **Plan inactive — read-only** until access is restored.
**Kimi K3 and Gemini 3.6 Flash.** Kimi K3 is available for long-horizon coding, agentic workflows, and large-context knowledge work. Gemini 3.6 Flash is available for fast coding, agentic work, and web development.
Switch workspaces, share with teammates, and understand Shared team memory.
Manage the team identity, members, plan, and shared-memory setting.
Compare membership-gated team sharing with public share links.
Choose between Kimi K3, Gemini 3.6 Flash, and other available models.
## New models: GPT-5.6, Grok 4.5, and Muse Spark 1.1
Three new frontier model families are available now in the Babbily model picker.
The coding results below are from [CursorBench 3.2](https://cursor.com/evals), Cursor's evaluation of agents on ambiguous, multi-file tasks from real Cursor sessions. Scores and average cost per task are a July 9, 2026 snapshot and may change as the evaluation is updated.
**GPT-5.6 Sol, Terra, and Luna.** GPT-5.6 Sol Max ranks third overall with a 67.2% score, behind Claude Fable 5 Max at 70.5% and Claude Fable 5 Extra High at 68.4%. GPT-5.6 Terra Max scores 64.9% at $2.73 per task, roughly half Sol Max's $5.22. Choose Luna when you want the faster, lower-cost GPT-5.6 option.
**Grok 4.5.** Grok 4.5 High scores 66.7% and ranks fourth overall, half a percentage point behind GPT-5.6 Sol Max. Its average cost is $1.51 per task compared with $5.22 for Sol Max—about 3.5 times less for a similar score.
Cursor marks Grok 4.5's CursorBench results with an asterisk because an earlier snapshot of the Cursor codebase was unintentionally included in its training data. Cursor says the score impact is unclear. Review the [CursorBench footnote](https://cursor.com/evals) before citing the result elsewhere.
**Muse Spark 1.1.** Use Meta's upgraded model for coding, debugging, long-running agentic tasks, and multimodal reasoning across supported inputs.
Choose a model based on your task, speed, context, tools, and API usage budget.
See Cursor's current scores, average per-task costs, methodology, and Grok 4.5 footnote.
## Get notified when Babbily is done
Babbily can now send a push notification when a chat, image, or video finishes running in the background.
## What's new
**Turn notifications on when prompted.** After your first prompt, Babbily may offer **Get notified when Babbily is done**. Choose **Allow notifications**, then allow notifications in your browser or device prompt.
**Manage notifications anytime.** Open **Settings > Account > Notifications** to turn **Chat completion notifications** on or off.
**Step away while work runs.** If you leave the tab or Babbily is running in the background, Babbily can let you know when the work is ready.
**Helpful for longer work.** Completion notifications are especially useful for longer generations, research-heavy chats, image creation, and video runs.
**Documentation is easier to find.** Settings now includes a direct **Documentation** link under **Learn more**.
Turn background completion notifications on or off.
Troubleshoot notification permissions and supported browsers.
## Claude Sonnet 5 is now available in Babbily
Claude Sonnet 5 is now available in the Babbily model picker. It is the recommended Sonnet model for deep reasoning, writing, coding, analysis, and complex multi-step work.
Sonnet 4.6 is still available, but it now appears as **Sunsetting** while teams move to Sonnet 5.
## Also added
**GLM 5.2 Fast mode.** GLM 5.2 now includes a Fast mode option when you want quicker responses for supported work.
**Nano Banana 2 Lite.** Use Nano Banana 2 Lite when you want lighter-weight image generation and faster iteration.
**Updated model ordering.** The model selector now places newer recommended models higher in the catalog.
Choose a model for reasoning, coding, speed, files, media, and API usage budget.
Create images and videos with supported models and tools.
## Sign in faster with passkeys
Babbily now supports passkeys for faster, passwordless sign-in on supported browsers and devices.
You can add a passkey using Face ID, Touch ID, Windows Hello, a password manager, or a hardware security key. Once added, you'll see **Sign in with passkey** on the web login screen.
New users may be prompted to add a passkey during setup, and existing users can add, rename, or remove passkeys anytime from **Settings > Account**.
Passkeys are currently available on web browsers. Native app passkey support will come later.
Learn how to sign in, add passkeys, and manage account access.
## News is faster, and Finance is now a dedicated workspace
Babbily now includes a full [Finance](/features/finance) area for market research, company pages, earnings, prediction markets, and watchlists. [News](/features/news) also has richer discovery, source-backed article generation, local context, and market highlights in the side rail.
## What's new
**Finance in Babbily.** Open **Finance** to review US markets, crypto, earnings, prediction markets, market summaries, company pages, price charts, key issues, analyst context, peers, and historical data where available.
**Watchlists and company research.** Add companies to a watchlist, open company pages from market rows, and ask follow-up questions from a Finance page with the page context already attached.
**Richer News discovery.** News now supports topic tabs, location-aware discovery, source citations, generated article views, article follow-up chat, weather, market outlook, and trending company links.
**Better performance and fallback behavior.** News and Finance data now refresh more smoothly, keep recent results available during temporary provider issues, and load common market and news panels faster.
Finance and market content in Babbily is informational. Verify important details with original sources and do not treat Babbily as financial advice.
Browse News, open sourced articles, update location, and continue from an article in chat.
Use market overviews, company pages, watchlists, earnings, prediction markets, and Finance research.
## Sakana Fugu Ultra, Kimi K2.7, and GLM 5.2 are now available
Babbily now includes three more models for advanced reasoning, coding, long-context work, and agentic task execution.
## What's new
**Sakana Fugu Ultra.** Use Fugu Ultra for complex prompts that benefit from multiple perspectives, deeper reasoning, and careful synthesis across approaches.
**Kimi K2.7.** Use Kimi K2.7 for coding-focused work, debugging, multi-file reasoning, long-context development, and agentic software tasks.
**GLM 5.2.** Use GLM 5.2 for long-horizon software engineering, repo-scale reasoning, refactoring support, and technical planning.
These models are available now in Babbily.
See the public release note for Sakana Fugu Ultra, Kimi K2.7, and GLM 5.2.
## New: Canvas Mode in Babbily 1.04
Babbily now includes [Canvas](/features/canvas-and-artifacts), a dedicated workspace for creating, editing, and refining AI-generated work without leaving chat.
## What's new
**Edit generated work in Canvas.** Drafts, tables, code, image briefs, and structured outputs can open in a durable side panel where you can keep working.
**Keep your edits in context.** When you continue prompting in the same thread, Babbily can use the latest Canvas artifact state instead of only the original chat reply.
**Render structured outputs.** Babbily can turn supported structured JSON outputs into cleaner visual components such as dashboards, tables, metrics, and callouts.
**Export or send work onward.** Supported artifacts can be downloaded as Word, PDF, CSV, source, or JSON files. Some artifacts can also be sent to connected apps when plugin actions are available.
See the public release note for Canvas Mode in Babbily 1.04.
## Claude Fable added to Babbily
Claude Fable 5 is now available on Babbily — with a 1 million token context window.
Fable is a Mythos-class model built for long-running, complex, and asynchronous work, with robust safeguards designed for serious tasks. That means fewer interruptions, less context loss, and more room to think through big projects from start to finish.
Available now on Babbily.
## New features
**Skills, tools, memory, auto mode, and plugins.** Babbily can now personalize replies with persistent [memory](/features/memory), use [tools and skills](/features/tools-and-skills) automatically when they help, and pull live context from [connected apps](/features/plugins). Auto mode can choose available capabilities when they seem useful.
**API usage billing.** A new monthly [API usage budget](/features/usage-budget) replaces fixed message limits. Track what you've used, see what counts, and know when it resets — all from the chat sidebar.
## Updates
**Clearer context indicator.** The [context window](/features/context-window) tracker now updates more reliably during long chats so you can tell when Babbily will compact older turns.
**Budget refresh.** Your usage view refreshes faster after each turn so recent activity is reflected quickly; final usage may continue to settle as billing data refreshes.
## Fixes
* Resolved a stall when the context-limit warning appeared mid-send.
* Restored model fallback so [chat](/features/chat-and-models) keeps working when a preferred model is briefly unavailable.
* Fixed scrolling and layout glitches in [temporary chats](/features/temporary-chat) and on tablet viewports.
* Sign-in via Google and Apple is now more reliable on mobile.
# Accessibility and supported environments
Source: https://docs.babbily.com/features/accessibility-and-supported-environments
Use Babbily across modern browsers and devices, and learn what to do if accessibility or environment issues appear.
Babbily is designed for modern web browsers and supported app surfaces. Some features depend on your browser, device, permissions, selected model, plan, and account state.
Troubleshoot microphone, upload, pop-up, plugin authorization, and device-specific issues.
## Supported environment basics
For the best experience:
* Use a current version of a modern browser.
* Enable JavaScript and required browser permissions.
* Allow pop-ups when connecting apps.
* Allow microphone access when using voice dictation.
* Keep your browser zoom and operating system display settings at comfortable levels.
* Refresh Babbily after changing browser permissions, plan state, connected apps, or account settings.
## Feature dependencies
| Feature | What it may depend on |
| -------------------------- | -------------------------------------------------------------------------------------------------------------- |
| Voice dictation | Browser support, microphone permission, device support, and account state. |
| File attachments | File type, file size, model support, browser upload behavior, and account state. |
| Plugins | Pop-up permissions, connected-app authorization, plugin readiness, and your role in the connected app. |
| Image and video generation | Selected model, usage budget, source media support, and plan/account state. |
| Sharing | Account state, shared thread visibility, included attachments or media, and public search visibility settings. |
## Accessibility tips
* Use browser zoom if text is too small.
* Use operating system contrast, motion, and display settings when helpful.
* Use keyboard navigation where supported by your browser and the active control.
* Add clear prompt structure when using screen readers, such as numbered tasks or requested headings.
* Prefer shorter chats or separate threads when a long page becomes difficult to navigate.
If your organization requires a formal accessibility review or specific compliance documentation, contact support with the requirement and timeline.
## Report an accessibility issue
When contacting support, include:
* The page or feature.
* Browser, device, and operating system.
* Assistive technology involved, if any.
* What you expected to happen.
* What happened instead.
* A redacted screenshot or short description of the issue.
Do not include passwords, verification codes, private keys, payment details, or sensitive files in the report.
# Account access
Source: https://docs.babbily.com/features/account-access
Create your account, verify your email, sign in with passkeys, reset your password, accept team invites, and manage account access.
Use this guide when you are joining Babbily for the first time, returning to your account, or helping a teammate get access.
Fix sign-in, verification, password reset, and team invitation issues.
Add passkeys for faster, passwordless sign-in on supported web browsers and devices.
Invite teammates, manage members, and understand owner-only team controls.
## Create an account
Go to Babbily and choose **Sign Up**.
Continue with a supported social sign-in option or enter your email address.
If Babbily sends a confirmation code, enter the code from your email.
Add your name, confirm the required account details, and complete the onboarding screens. New users may be prompted to add a passkey during setup.
After setup, open a new chat and send your first message.
Babbily may ask you to confirm that you meet the minimum age requirement before continuing.
## Sign in
Use the same method you used to create your account. If you signed up with a social sign-in option, choose that option again. If you created a password for email sign-in, use your email address and password.
If you have added a passkey, choose **Sign in with passkey** on the web login screen and follow your browser or device prompt.
If Babbily says the account was not found, check that you entered the same email address you used during sign-up. If Babbily says the account already exists, switch to **Log In** instead of **Sign Up**.
## Use passkeys
Passkeys let you sign in without typing your password on supported web browsers and devices. You can use Face ID, Touch ID, Windows Hello, a password manager, or a hardware security key when your browser and device support it.
### Add a passkey
Go to **Settings > Account**.
In **Passkeys**, choose **Add passkey**.
Complete the browser, device, password manager, or security key prompt.
Name the passkey if prompted so you can recognize it later.
New users may also see **Set up passkey** during setup. If you skip it, you can add one later from **Settings > Account**.
### Manage passkeys
Open **Settings > Account** to add another passkey, rename a passkey, or remove a passkey you no longer use. A setup reminder may stop appearing once you have a registered passkey; you can still manage passkeys from settings.
Passkeys are currently available on web browsers. Use another sign-in method in the native app if passkeys are unavailable. If **Sign in with passkey** or **Add passkey** is hidden, your current browser, device, or account state may not support passkeys yet.
## Reset your password
If you use email and password sign-in:
Choose **Forgot your password?** from the sign-in page.
Use the email address on your Babbily account.
Enter the reset code Babbily sends you.
Choose a new password and return to sign-in when Babbily confirms the update.
If you signed up with a social sign-in option and did not add a password, continue with that sign-in option or use the account settings flow when available.
## Accept a team invite
Team invitations are sent by email. Open the invitation link from the account you want to use with the team.
Use the latest invite email from your team owner.
If prompted, sign in. New users may need to finish onboarding before accepting the invite.
Review the team name, then choose **Accept invitation**.
After accepting, open Babbily and check **Settings > Team**.
If you use multiple email addresses, ask the team owner which email they invited before creating a new account.
## Manage account settings
Open **Settings > Account** to review account details, update available profile fields, manage passkeys, request password help, or log out. For data you cannot remove yourself, account deletion requests, or access issues that remain after troubleshooting, contact support.
Understand what you can copy, download, delete, or request through support.
# Account and settings
Source: https://docs.babbily.com/features/account-and-settings
Find profile, appearance, privacy, notifications, billing, team, plugin, skill, and memory settings.
Use **Settings** to manage your Babbily account, workspace preferences, and connected features.
## Open settings
Select your account menu, then choose **Settings**.
Available tabs can depend on your account, plan, team role, and device.
## Switch workspaces
The workspace list appears in your account menu, and the sidebar shows a workspace pill for the active workspace.
Select the workspace pill or open your account menu.
Select **Personal** or a team workspace. Babbily opens a clean chat surface in that workspace.
The workspace list shows the plan context for Personal and each team. An inactive team is labeled **Plan inactive — read-only**.
Chats, Library items, and memory remain with the workspace where they were created. Switching workspaces does not move existing work.
If you do not have a team workspace, the list can show **Create a team workspace**.
## Settings tabs
| Tab | What you can manage |
| ---------- | ------------------------------------------------------------------------------- |
| Profile | Name, nickname, location, and profile details. |
| Appearance | Display preferences. |
| Account | Account details, passkeys, notifications, password help, and sign-out controls. |
| Privacy | Privacy-related preferences and account help. |
| Billing | Plan, usage, renewal, and billing controls. |
| Team | Team identity, members, invitations, Shared team memory, and team status. |
| Plugins | Connected apps and authorization status. |
| Skills | Verified skills and your markdown skills. |
| Memory | Memory Profile, saved context, citations, and memory controls. |
## Common tasks
Review your Memory Profile and delete saved context.
Authorize apps Babbily can use in chat.
See your API usage budget and plan limits.
Add, rename, or remove passkeys from **Settings > Account**.
Turn background completion notifications on or off.
Invite members and manage team billing.
## Manage notifications
Chat completion notifications let Babbily notify you when a chat, image, or video finishes while Babbily is in the background.
You can turn them on after your first prompt or anytime from settings.
After your first prompt, Babbily may show **Get notified when Babbily is done**. Choose **Allow notifications**, then allow notifications in your browser or device prompt.
Go to **Settings > Account > Notifications**.
Switch **Chat completion notifications** on when you want Babbily to notify you about background runs. Switch it off anytime to stop completion notifications.
Keep the chat, image, or video run going. If Babbily finishes while the tab is inactive or in the background, it can notify you that the work is ready.
Notifications are optional. Your browser or device may also need to allow notifications for Babbily before completion notifications can appear.
### Notifications on iPhone and iPad
For web notifications, add Babbily to your Home Screen first: open the browser's **Share** menu, choose **Add to Home Screen**, then open Babbily from that icon. Sign in if needed and enable **Chat completion notifications** there.
If notifications are blocked, change the browser or device permission before trying again. The first-prompt offer can be dismissed without interrupting your chat.
## Open documentation from settings
To open this help center from Babbily, open your account menu, choose **Learn more**, then choose **Documentation**.
## When a setting is missing
A setting may be hidden when:
* You are signed out.
* Your plan does not include that feature.
* You are a team member and the control belongs to the team owner.
* The feature is unavailable on your current device or browser.
* Your browser or device has blocked a required permission, such as notifications.
If you recently changed plans, joined a team, or connected an app, refresh settings after a moment.
# Admin and procurement FAQ
Source: https://docs.babbily.com/features/admin-and-procurement-faq
Answer team, billing, SSO, privacy, support, plugin, and access questions before buying or expanding Babbily.
Use this FAQ when you are evaluating Babbily for a team, preparing an internal purchase, or answering admin questions.
## Team and access
Team owners manage the workspace identity, invitations, member removal, Shared team memory, plan changes, billing, and team usage controls where available. Team members can use the team plan but may not see owner-only controls.
Yes. Every account has a Personal workspace, and accepted team memberships add team workspaces to the account menu. Chats, Library items, and memory stay with the workspace where they were created.
Team setup requires at least three people total, including the owner. Pending invitations do not receive team plan access until accepted, but may affect billed seats until accepted, declined, or canceled.
Removed members stop receiving team plan access and lose membership-gated access to team shares. Removing a member does not delete content inside that person's Personal workspace or connected apps.
Team plan cards may list SSO integration. Availability and setup can depend on the plan and account, so contact support if SSO is a procurement requirement.
Team plan cards may list team analytics or usage tracking. Exact visibility can depend on your plan, account state, and role.
## Billing and support
Team plans use seat-based billing with a three-person minimum, including the owner. Team owners manage subscription and billing controls. Team members may have limited billing visibility.
Yes. Individual and team plans can coexist, and joining or creating a team does not automatically cancel an individual subscription. A team owner switches to the team workspace they own before changing that team's plan.
Accepted team members receive the team plan's included API usage budget and follow the team plan's reset timing.
Some paid and team plan cards list priority support. Team plan cards may also mention a dedicated account manager. Contact support if you need to confirm support coverage for your account.
Include the plan you are considering, estimated number of accepted users, billing owner email, SSO needs, support expectations, privacy questions, and any required timeline.
## Data, privacy, and plugins
Open **Settings > Privacy** or visit the Babbily Legal Zone linked from the product. The privacy page summarizes how Babbily handles conversations, data selling, deletion requests, safety, account communication, and optional features.
Babbily's privacy settings state that Babbily does not use your chats to train generative models by default unless you have given explicit permission.
Plugin access belongs to the account that authorized it and is also limited by the connected app's own permissions. Workspace admins should manage app permissions from the connected app when needed.
The team owner can enable Shared team memory from **Settings > Team**. When enabled, useful facts from team-workspace chats can become shared context for current members. Personal Memory Profiles remain separate. Decide what team-wide context is approved before enabling the setting.
Yes. Supported team-workspace chats and saved media can be shared read-only with specific teammates or everyone in the team. Access depends on current membership and is separate from public share links.
Team membership changes do not automatically delete connected-app content or permissions. Ask affected users to review **Settings > Plugins**, and use the connected app's admin settings when your organization needs to remove app access.
Do not enter passwords, verification codes, API keys, private keys, payment card numbers, government identifiers, raw connected-app credentials, or confidential data you are not allowed to process in Babbily.
Switch workspaces, share with teammates, and understand Shared team memory.
Manage identity, roles, invitations, memory, billing, usage, and troubleshooting.
Understand plan changes, resets, and team billing.
Review practical data controls for everyday use.
# Billing and plan changes FAQ
Source: https://docs.babbily.com/features/billing-and-plan-changes-faq
Understand upgrades, downgrades, resets, team billing, and missing billing controls.
Use this FAQ when you are changing plans, reviewing usage, or helping a team understand billing visibility.
## FAQ
Open **Settings > Billing**. Depending on your account, role, and plan, you may see your current plan, usage, renewal timing, upgrade or downgrade actions, cancel controls, or a billing portal link.
Upgrading can increase your monthly API usage budget and unlock plan-specific features. Some changes may apply quickly, while others can take a short time to appear across usage and billing surfaces.
Downgrades or cancellations can reduce future usage, change renewal timing, or remove plan-specific features. Some changes may be scheduled for the next billing period.
Individual plans reset on the usage period shown in your account. Paid plans usually align with the billing period. Team members follow the timing shown for the team plan.
No. Unused API usage budget does not roll over to the next period.
Team plans use seat-based billing with a three-person minimum, including the owner. Accepted members receive team plan access and API usage budget. Pending invitations do not receive access or usage, but may affect the billed seat count until accepted, declined, or canceled.
Yes. Individual and team plans can coexist. Joining or creating a team does not automatically cancel your individual subscription, and your Personal workspace remains separate from the team workspace.
Team plan changes are scoped to the team workspace the owner manages. Switch to a team workspace you own, then return to the team plan or billing controls. This prevents a change intended for one plan from affecting another workspace.
The team workspace is still visible, but new work is locked because its team plan is not active. The team owner should restore or complete the team subscription before members continue working there.
Billing controls can depend on your role. Team members may have limited billing visibility and should ask the team owner to manage plan or subscription changes.
Usage and billing cards can take a short time to refresh after model work, media generation, team membership changes, or plan updates. Refresh settings and check again after a short wait.
Team plan cards may list shared workspace features, centralized billing and user management, team analytics or usage visibility, SSO integration, and priority support. Availability and setup can depend on your plan and account, so contact support if you need help confirming a specific requirement.
Include your account email, plan, whether you are a team owner or member, what you expected, what happened instead, and the approximate date, time, and timezone. Do not send full payment card numbers or sensitive credentials.
Review plan options, team billing, and plan-change basics.
See included usage amounts, what counts, and reset timing.
Review team, SSO, privacy, support, and procurement questions.
Learn what to include in a support request.
# Billing and plans
Source: https://docs.babbily.com/features/billing-and-plans
Manage your Babbily plan, billing details, renewals, and team subscriptions.
Babbily plans control your monthly API usage budget, billing cycle, support level, and team features.
## Plan options
| Plan | Best for |
| ----------- | -------------------------------------------------------------------------------- |
| Free | Trying Babbily and lightweight usage. |
| Personal | Individual work with a larger monthly budget. |
| Premium | Individual work with more usage and priority benefits. |
| Ultra | High-usage individual workflows, including more frequent media or research work. |
| Team | Shared workspace and billing for teams. |
| Teams Ultra | Higher team usage for larger or heavier workflows. |
Scheduled Tasks are available on paid plans. Each run uses the API usage budget of the personal or team workspace where the task is saved. A workspace can contain up to 10 personal Scheduled Tasks or 25 team Scheduled Tasks.
## Keep individual and team plans
Individual and team plans can coexist on the same account. Joining or creating a team does not automatically cancel an individual subscription.
* Your individual plan remains associated with your **Personal** workspace.
* A team plan is associated with the team workspace.
* The account menu shows the plan context beside each available workspace.
* Billing changes for one plan family do not automatically replace the other.
A chat remains tied to the workspace where it was created. Switching workspaces does not move the chat or its earlier usage.
## Open billing settings
Go to **Settings > Billing** to review your current plan, usage, renewal timing, and billing controls.
Depending on your account and plan, you may be able to:
* Upgrade.
* Downgrade.
* Cancel.
* Open the billing portal.
* Manage team subscription details.
## Team billing
Team plans use seat-based billing with a three-person minimum, including the owner. Accepted members receive team plan access and API usage budget. Pending invitations do not receive access or usage, but may affect the billed seat count until accepted, declined, or canceled.
Team owners manage the subscription. Team members can use their team entitlement but may not see owner-only billing controls.
Team plan cards may include shared workspace features, centralized billing and user management, team analytics or usage visibility, SSO integration, and priority support. Availability and setup can depend on the plan and account.
If you own a team, switch to that team workspace before changing its plan. A team workspace with an inactive plan is labeled **Plan inactive — read-only** until access is restored.
## Changing plans
Plan changes can affect:
* Available monthly usage.
* Renewal or reset timing.
* Team member usage.
* Access to plan-specific features.
Some changes apply immediately. Others may be scheduled for the next billing period.
Check [API usage budget](/features/usage-budget) before heavy research or media generation so you know how much room remains in the current period.
Review paid-plan access, workspace limits, and usage behavior.
Understand plan changes, refresh timing, missing billing controls, and team billing.
Review team, SSO, support, privacy, and procurement questions.
# Built-in tools reference
Source: https://docs.babbily.com/features/built-in-tools-reference
Learn when to use Babbily's built-in tools for search, research, site maps, weather, QR codes, media, and more.
Babbily includes built-in tools that help a chat do more than answer from the selected model alone. Use this reference when you want to choose a tool manually or understand why Babbily used one in **Auto** mode.
Tool availability can vary by model, plan, remaining API usage budget, account state, and tool mode. If a tool is hidden, try a supported model, switch from **Off** to **Auto** or **Manual**, and check your usage.
## Choose a tool mode
| Mode | What it does | Use it when |
| ---------- | --------------------------------------------------------------- | --------------------------------------------------------------------- |
| **Auto** | Babbily decides which available tools fit your request. | You want the fastest path and do not need to force a tool. |
| **Manual** | You choose the exact tool, plugin, or skill for the next reply. | You know the answer needs search, research, media, or a specific app. |
| **Off** | Babbily answers without tools. | You want a model-only draft, brainstorm, rewrite, or quick answer. |
## Built-in tools
Use web search for focused questions that need current public sources.
Good uses include vendor pricing, product comparisons, recent news, current policies, public facts, and source-backed summaries.
Try:
```text theme={null}
Search the web for recent public announcements from these companies and cite sources.
```
Search can add up to three pictures above the answer when the subject is something you can see or when you ask to see it, for example `Show me photos of the new terminal building.` Follow-up questions get pictures only when you ask, and Babbily leaves out logos, icons, ads, and unrelated pictures. See [how to ask for pictures from web search](/features/visual-replies-in-chat#ask-for-pictures-from-web-search).
Search can fail or return limited results if the topic is hard to find, access is blocked, or the selected model does not support the tool. Verify high-stakes answers with authoritative sources.
Use deep research when you need a broader, multi-source report instead of a quick lookup. The Deep Research Agent plans sub-questions, opens the report outline in Canvas, gathers evidence, and streams a cited report.
Good uses include competitive briefs, policy research, vendor comparisons, market landscapes, and strongest-argument summaries.
Try:
```text theme={null}
Do deep research on this market and return a sourced brief with key players, risks, and open questions.
```
Deep research can take several minutes and may use more API usage budget. If evidence gathering reaches its time limit, Babbily finishes from the available sources and tells you coverage is partial. If writing reaches its limit, Canvas keeps the written portion and labels it incomplete. Reports cite text sources and do not include search pictures.
Use site map to inventory, review, or summarize public pages on a website.
Good uses include finding product pages, checking documentation coverage, auditing public links, or discovering what a public site contains.
Try:
```text theme={null}
Map this public website and summarize the main product, pricing, and docs pages.
```
Site map reviews website structure. It does not map geographic places. Private, login-only, blocked, or heavily dynamic pages may be incomplete.
Use finance research in chat for company, market, earnings, filings, peer, or financial-topic research.
Good uses include public company summaries, earnings context, market comparisons, historical context, and financial source review. For the dedicated Finance workspace, see [Finance](/features/finance).
In a supported chat started from Finance, **Auto** uses Finance research automatically. **Manual** starts with the tool enabled for that Finance thread, and **Off** disables tools. The Finance context and setting stay with follow-ups and **Retry**.
Try:
```text theme={null}
Research this public company and summarize recent earnings, major risks, and peer context.
```
Finance research checks the available source coverage and may use primary filings to complete historical comparisons. If a company does not report a requested metric separately, Babbily should explain the limitation instead of guessing. Finance research is informational and should not be treated as financial advice.
Use weather for current conditions and short forecasts by location.
Good uses include travel planning, event planning, quick local checks, and comparing forecast windows.
Try:
```text theme={null}
Show the 5-day forecast for Denver in Fahrenheit.
```
Weather results depend on whether the location can be resolved and current public weather data is available. Do not rely on Babbily for safety-critical weather decisions.
Use stats for supported compact public-data stat cards.
Good uses include supported public-data questions such as earthquake counts by date range and magnitude.
Try:
```text theme={null}
Show earthquake counts for the last 7 days by magnitude range.
```
For your own KPIs, attach a file, paste the numbers and ask for a visual summary, or use a connected spreadsheet app when available.
Use QR code to create a scannable image from a public link or non-sensitive text.
Good uses include event links, public landing pages, signup forms, and simple labels.
Try:
```text theme={null}
Create a QR code for https://babbily.com.
```
Anyone who can scan the QR code can read the encoded content. Do not put passwords, verification codes, private keys, payment details, or confidential information in a QR code.
Use the **Image generation** tool to create a new image from a text prompt with **GPT Image 2.5 Flare**. To edit an attached image, select a supported image model directly. See [image model and tool differences](/features/image-and-video-generation#choose-the-image-model-or-chat-tool).
Good prompts describe subject, style, composition, lighting, aspect ratio, and constraints.
Try:
```text theme={null}
Create a landscape illustration of a collaborative AI workspace, bright office lighting, minimal background.
```
Image generation can take longer than text replies and usually uses more API usage budget. Review generated media before using it publicly.
Use video generation to create, animate, or edit short videos when your selected model and plan support it.
Good prompts describe the subject, movement, pacing, duration, and source media when supported.
Try:
```text theme={null}
Create a short video of a dashboard animating from an empty state to a completed weekly report.
```
Video generation can take minutes, uses more API usage budget than many text replies, and may vary by model, file type, duration, and account state.
## Get help with Babbily
Ask a product question in **Auto** mode, such as `How do I add an existing chat to a project?` Babbily can look up help-center pages and link the relevant instructions. When available, select **Babbily Docs** from the composer or use `/babbily-docs` to request this lookup directly.
For previous conversations, Babbily can use [past-chat lookup](/features/memory#refer-to-past-chats) in **Auto** mode when memory and the current chat support it. This is separate from searching public websites.
## Why a tool may be missing
Try these checks:
* Switch to a model that supports the tool.
* Change tool mode from **Off** to **Auto** or **Manual**.
* Check your remaining API usage budget.
* Remove conflicting file types or unsupported source media.
* Reconnect an app if the missing option is a plugin.
Learn how tools, skills, slash commands, and manual mode work together.
See how Babbily can show maps, charts, weather, QR codes, media, and citations.
# Canvas and artifacts
Source: https://docs.babbily.com/features/canvas-and-artifacts
Create, edit, save, export, and send AI-generated work without leaving chat.
Canvas gives you a dedicated workspace beside chat. When a reply becomes something you may want to keep working on, Babbily can open it as an artifact in the Canvas panel.
Use Canvas for drafts, tables, code, image briefs, and structured outputs that need more than a one-time chat reply.
Learn how Canvas fits into regular chat threads.
Find saved chats, media, translations, and Canvas artifacts.
## When Canvas appears
Babbily may open Canvas when you ask for substantial work that benefits from a separate editing surface.
Canvas artifacts can include:
* Text drafts and documents.
* Deep Research plans and cited reports.
* Code snippets and scripts.
* Image prompts or creative briefs.
* Tables and spreadsheet-style data.
* Structured visual outputs, such as dashboards, metrics, tables, callouts, and supported interactive views.
If a response includes an artifact card, open it to continue working in Canvas.
## Review a Deep Research report
When you use **Deep research**, Canvas opens after the research plan is ready. The artifact can show:
* The questions the agent is working through.
* The planned report sections.
* Research progress and sources found so far.
* The cited report as it is written.
The artifact card updates when the run finishes. If evidence gathering reaches its time limit, Babbily finishes from the available sources and notes in chat that coverage is partial. If writing reaches its limit, Canvas keeps the written portion and labels the report incomplete.
If a selected plugin action needs approval, the report pauses with the work gathered so far. Review the approval card inline in the conversation. If you allow the action, Babbily continues the same report. If you cancel, the paused run ends and Canvas keeps the sources gathered before the pause.
Learn when to use the Deep Research Agent, how to follow its plan, and how to review sources.
## Edit directly in Canvas
Canvas lets you edit supported artifacts without copying them into another editor first.
Depending on the artifact type, you can:
* Edit text or markdown.
* Adjust table-like content.
* Review code in a focused editor.
* Rename the artifact.
* Save a new version.
* Resize the side panel on desktop.
* Close the panel and reopen the artifact later.
Save your changes before switching away if you want Babbily to keep that version.
## Revise with the latest Canvas state
When you continue prompting in the same thread, Babbily can use the active Canvas artifact as context. That helps follow-up requests apply to the current version, including edits you made in the panel.
Good follow-up prompts include:
* `Make this shorter and keep the same section order.`
* `Turn the second section into a table.`
* `Update the code to accept a CSV file.`
* `Add a warning callout under the pricing notes.`
If you want to change topics, start a new thread or close the artifact so old Canvas context does not steer the next request.
## Use versions
Saving changes creates a new artifact version when version history is available. Use the version controls to move between earlier and later versions.
Versions are useful when you want to compare a draft before and after a prompt, recover older wording, or review how a table changed.
## Export and download
Export options depend on the artifact type.
| Artifact type | Available formats |
| ----------------- | ------------------------- |
| Text | PDF or Word document. |
| Sheet | CSV for spreadsheet apps. |
| Code | Source file download. |
| Structured render | JSON. |
Review exported files before sharing them outside Babbily.
## Send artifacts to connected apps
Some Canvas artifacts can be sent to connected apps from the artifact actions menu. Available destinations depend on your connected apps, permissions, and whether the plugin tools are ready.
If a destination is not connected, Babbily may ask you to connect it first. If an action writes to another app, review what will be sent before approving it.
Connect workspace apps and review approval-required actions.
## Run code artifacts
When a code artifact supports execution, use the run action in Canvas to test it. Code execution is intended for quick checks, not production workloads.
Execution support can vary by language and artifact state. JavaScript, TypeScript, JSX, TSX, and Python-like code can run through supported JavaScript or Python execution paths. Network access is unavailable during execution, and long-running code may stop before it finishes.
Review code before running it. Do not include passwords, API keys, private keys, customer data, or other secrets in code artifacts.
## Limits to expect
Canvas availability can vary by model, plan, tool mode, and the type of work you request.
Keep these expectations in mind:
* Babbily may answer in chat instead of Canvas when the request is simple.
* Structured visual artifacts may take a moment to prepare before the panel opens.
* Some generated structures are for viewing and editing, not for taking external actions.
* Export and plugin actions vary by artifact type.
* Plugin actions depend on your app permissions and plugin readiness.
If Canvas does not appear, ask directly for a draft, table, code artifact, or structured output. You can also try a model with stronger tool or file support.
# Chat and models
Source: https://docs.babbily.com/features/chat-and-models
Choose a model, send better prompts, and understand how model capabilities affect your chat.
Chat is the center of Babbily. You can ask questions, draft content, analyze files, research topics, create media, and continue work across a thread.
Choose a model for speed, reasoning, files, tools, media, context, and API usage budget.
Create, edit, save, export, and send generated work from a dedicated Canvas.
Prefer speaking to typing? Use voice dictation in chat or Translate when your device supports it.
## Choose a model
Use the model picker to choose the model for a chat. Models can differ in:
* Speed.
* Reasoning depth.
* Writing style.
* Context window size.
* File support.
* Tool and media support.
If a tool or file type is hidden, the selected model may not support it. Switch models if you need a capability that is not available.
Some models also show extra selector cues:
* **Fast mode** means the selected model can use a faster response path for supported work.
* **Sunsetting** means the model is still available, but Babbily recommends a newer option for new chats when one is available.
Recent additions include **GPT-6 Astra**, **Gemini 3.8 Flash**, **GPT Image 2.5 Flare**, **Claude Fable 5.1**, and **GLM 5.3 Flash**. Some older models remain selectable as **Sunsetting**; others have been removed from the picker. See [current model transitions and alternatives](/features/model-capabilities#sunsetting-and-removed-models) before continuing work with an older model.
## Adjust reasoning
Choose **Manual** tool mode and turn on **Reasoning**. Models with adjustable effort show a **Reasoning effort** selector beside the microphone. Other reasoning-capable models offer an on/off toggle without a level selector.
See [reasoning controls by model](/features/model-capabilities#choose-reasoning-effort) for the September additions. Reasoning effort and **Fast mode** are separate controls.
## Use Fast mode
Fast mode can be useful for drafting, brainstorming, coding, and iterative work where speed matters.
Open the model picker and choose a model with the Fast mode lightning icon. Claude Opus 5 supports Fast mode.
Choose **Fast mode** using the lightning control near the composer.
Babbily uses the faster response path while keeping your selected model visible in the chat.
Fast mode can use more of your API usage budget. Babbily turns it off when you change chats or select another model, so turn it on again when you still want the faster path.
The **Fast mode** control appears only for supported models. “Flash” in a model name does not mean it supports this toggle. Check [Fast mode availability](/features/model-capabilities#use-fast-mode) if the control is missing.
## Write better prompts
Good prompts usually include:
* The outcome you want.
* The audience.
* Any constraints or examples.
* The format you want back.
* Whether Babbily should use search, files, memory, or connected apps.
```text theme={null}
Draft a 6-bullet launch update for the product team. Use a calm, direct tone. Include risks and next steps.
```
## Continue a thread
Follow-up messages can refer to earlier replies, files, and context in the same thread. If a conversation gets long, Babbily tracks the context window and may compact older context so the thread can continue.
## Edit an earlier message
Use **Edit message** on one of your own messages, revise the text, and choose **Send**. Babbily generates a new reply using the selected model and current generation controls.
Sending an edit replaces that message and removes all later messages in the thread. Copy any later work you want to keep before sending.
Wait for the thread's current action to finish before editing. Original attachments must still be available and supported by the selected model. If an attachment is missing, send a new message with the file attached. If Babbily says the edit may have been saved, refresh the thread before trying again.
## Refer to an earlier chat
In **Auto** mode, Babbily can look up eligible earlier chats when you ask about a previous discussion. This requires memory to be enabled and a model that supports it. It is a limited search, so give a topic or other identifying detail and review any linked conversation.
See [past-chat lookup and its boundaries](/features/memory#refer-to-past-chats).
## Get notified when a run finishes
For longer chats or research-heavy prompts, Babbily can notify you when a background run finishes.
After your first prompt, Babbily may offer **Get notified when Babbily is done**. Choose **Allow notifications** if you want completion notifications, then allow notifications in your browser or device prompt.
You can also turn **Chat completion notifications** on or off anytime in **Settings > Account > Notifications**.
Completion notifications are optional and depend on browser or device notification support.
## Work in Canvas
When a reply becomes a document, table, code snippet, image brief, or structured output, Babbily may open it as a Canvas artifact. Canvas gives you a side workspace where you can edit, save versions, export files, and continue prompting with the latest artifact state included as context.
Use Canvas when you want to refine generated work instead of copying it into another app first.
## Use chat history
Saved chats appear in the sidebar and Library. You can reopen, rename, share, or delete chats from the available thread actions. You can also [add or move your own eligible threads into a project](/features/projects#add-an-existing-thread).
## When to start a new thread
Start a new thread when:
* You switch topics.
* Earlier context is no longer useful.
* The context indicator is near the limit.
* You want a cleaner answer without old assumptions.
If Babbily keeps using old context, start a new thread and restate only the details that still matter.
Write clearer prompts with goals, context, audience, constraints, and output format.
Understand what can vary by model, tool, file, plugin, plan, usage, and context.
# Context window
Source: https://docs.babbily.com/features/context-window
Understand the context indicator and how Babbily keeps long chats usable.
Every model has a context window. This is the amount of conversation, file content, and tool output the model can consider for one reply.
Babbily shows a context indicator near the composer so you can see how full the current chat is.
## What the indicator means
The context indicator estimates how much of the active model's window your chat is using. The details are approximate because models count conversation, file, and tool context differently.
* **Used** shows the percentage consumed.
* **Usage** shows the estimated context used out of the model's window.
* **Remaining** shows the approximate space left for the next message and reply.
* **Auto compact** shows when Babbily may summarize older parts of the chat.
The indicator updates as you send messages, attach files, receive replies, and switch models.
## Automatic compaction
In saved chats, when a long thread approaches the selected model's limit, Babbily can compact older conversation history into a shorter summary. Your visible chat history remains available, and the current conversation can keep going. A summary may omit details, so restate exact requirements when they matter.
Compaction helps:
* Preserve recent messages in detail.
* Keep the thread within the active model's limit.
* Continue long-running work without forcing a new chat immediately.
A compacted summary is still part of the model's context. Your usage percentage may drop, but it will not reset to zero.
Temporary Chats do not use automatic compaction. If they reach the limit, start a saved thread to continue.
## If context preparation fails
If Babbily cannot safely prepare the conversation for another reply, it may stop and ask you to try again. Refresh the thread before retrying. If the issue persists, start a new saved chat with the essential context rather than assuming all earlier details carried over.
Compaction is separate from saved memory: it summarizes a thread for continued work and does not mean its contents were added to your Memory Profile.
## Switching models
Different models support different context sizes. If you switch to a smaller model, Babbily recalculates the thread against that model's limit. If needed, Babbily may compact the thread before the next reply.
## When a chat gets too full
If a thread reaches the limit, Babbily may ask you to start a new thread. Memory and plugins can still help you carry important context forward.
Long files, pasted code, and many follow-ups about a Canvas or research report can fill context quickly. Start a new thread when your topic changes or when you no longer need earlier details.
# Data controls and privacy FAQ
Source: https://docs.babbily.com/features/data-controls-and-privacy-faq
Review practical controls for memory, temporary chats, plugins, sharing, files, Library items, and account help.
This FAQ explains the customer-visible controls you can use to manage what you share with Babbily.
Babbily does not use your chats to train generative models by default unless you give explicit permission, and Babbily does not sell your personal information. Babbily may use account information for verification, billing, support, safety, product communications, and aggregated usage insights.
Do not send passwords, verification codes, private keys, payment card numbers, government identifiers, or confidential files you are not allowed to use in Babbily.
## Quick control map
| Area | Where to manage it | What you can do |
| ------------------ | ------------------------- | -------------------------------------------------------------------------------------------- |
| Personal memory | **Settings > Memory** | Review your Memory Profile, remove eligible citations, or ask Babbily to forget something. |
| Shared team memory | **Settings > Team** | See whether team-wide memory is enabled. Team owners can turn it on or off. |
| Plugins | **Settings > Plugins** | Connect apps, review authorization status, and manage access where controls are available. |
| Temporary Chat | Composer or chat controls | Start a chat that is not saved like normal chat history. |
| Sharing | Chat share controls | Create, review, make private, or stop using share links. |
| Library | **Library** | Review saved chats, translations, and media. Delete supported saved items. |
| Account | **Settings > Account** | Update account details, manage passkeys, request password help, log out, or contact support. |
## FAQ
You control what messages you send, what files you attach, what memory you keep, which apps you connect, which plugin actions you approve, which chats you share, and which supported Library items you delete.
Open **Settings > Memory** to review your Memory Profile and memory citations. You can ask Babbily to forget something, or delete eligible citations from the Memory page.
Try:
```text theme={null}
Forget that I prefer long answers.
```
After a confirmed removal, the item is no longer used as saved memory. Original chat history remains and may still supply that information in the current conversation or eligible past-chat lookup. Remove that history separately if needed. See [memory controls and review](/features/memory). If you cannot remove something from the UI, contact support.
No. Your personal Memory Profile remains separate. A team owner can enable **Shared team memory** for a team workspace, which lets useful facts from team-workspace chats become shared context for current members.
All members can see whether the setting is enabled. Authors and the owner can delete eligible shared entries, and turning the setting off stops new additions.
Temporary Chat is for conversations you do not want saved in normal chat history or Library. Temporary chats are not meant for sharing, do not use or update Babbily memory, and are not used to train Babbily's models.
Temporary Chat is still a work surface, so avoid entering passwords, private keys, payment details, or other secrets.
Plugins let Babbily use access you authorize in the connected app. Connected apps may limit what Babbily can see or do based on your role and permissions.
Some connected-app actions ask for approval before they run. Review sends, updates, billing-related changes, customer-data actions, and workspace changes carefully before approving.
Use **Settings > Plugins** to manage connected apps where controls are available. If you do not see a disconnect option, revoke access from the connected app's own settings and contact support if you still need help.
Removing plugin access does not delete content inside the connected app.
When you make a chat public, anyone with the share link can view the messages plus any attachments, generated media, citations, or other content included in the shared thread. They do not get access to your account, settings, billing, plugins, or memory controls.
Public shared threads may be discoverable in Google Search Results when public search visibility is enabled. Make a link private when the thread no longer needs to be public.
Review the thread before sharing it.
Team sharing is available for supported team-workspace chats and saved media. You can share with specific teammates or everyone in the team, and recipients get read-only access under **Shared with you** in Library. Access ends when they leave the team.
Team sharing does not make the item public. Public links use separate visibility controls.
You can delete supported Library items such as chats, saved media, and translations. Deleting a saved media item removes it from your Library. Deleting a chat removes the chat from your saved thread list.
Some deletion actions cannot be undone.
In **Settings > Account**, you can update account details, manage passkeys, request password help, or log out. For account deletion or data you cannot remove yourself, contact support. Some information may need to be retained where required for safety review or feedback handling.
## Before contacting support
Include:
* Your account email.
* The page or feature involved.
* What you expected to happen.
* What happened instead.
* Approximate date, time, and timezone.
* Browser, device, and operating system.
* Your plan or team role if it matters.
* Error text or screenshots with sensitive details hidden.
Learn the practical privacy basics for everyday use.
See what to include and what not to send when you need help.
# Data export and deletion
Source: https://docs.babbily.com/features/data-export-and-deletion
Understand what you can copy, download, delete, disconnect, or request through support.
You control what you put into Babbily and can remove many items from the product UI. Some account-level requests or records may require support.
Understand practical ways to control what you share with Babbily.
Review controls for memory, Temporary Chat, plugins, sharing, Library items, and account help.
## What you can usually save or export
| Content | Customer action |
| ---------------- | ---------------------------------------------------------------------------------------- |
| Chat text | Copy important messages before deleting or leaving a temporary workflow. |
| Translations | Copy the translated text or continue the translation in chat when available. |
| Generated images | Save or download from the image controls when available. |
| Generated videos | Save, preview, fullscreen, or download when controls are available. |
| Shared work | Copy the share link when sharing is enabled. Review the shared thread before sending it. |
| Source citations | Open citations and save source links separately when needed. |
Available actions can vary by content type, model, plan, browser, and account state.
## What you can usually delete or disconnect
* Delete supported saved items from **Library**.
* Delete or update memory from **Settings > Memory** when controls are available.
* Ask Babbily to forget eligible saved context in chat.
* Disconnect apps from **Settings > Plugins** when controls are available.
* Revoke connected-app access from the connected app's own settings.
* Delete shared links or saved work where the product exposes a delete action.
Deleting saved memory and deleting a conversation are separate actions. After you forget a memory, the original messages may remain visible or searchable. Review both **Settings > Memory** and the original chat when removing information. See [memory and past-chat lookup](/features/memory#refer-to-past-chats).
## Account deletion and support requests
For account deletion or data you cannot remove yourself, contact support. Include the account email, the type of data involved, what you already tried, and whether the request is for an individual account or a team workspace.
Some information may need to be retained where required for billing, security, legal, support, abuse-prevention, or operational reasons.
## Before deleting
Copy or download important output before deleting a chat, file, media item, or Library entry.
Check whether a chat or item has been shared with others.
Open **Settings > Memory** and remove context you no longer want Babbily to reuse.
Disconnect or revoke connected-app access you no longer want to use.
Ask support for help with account-level deletion or anything not available in the UI.
Deletion can be permanent. Save copies of work you need before removing items.
# Deep research and search
Source: https://docs.babbily.com/features/deep-research-and-search
Use web search for focused lookups or the Deep Research Agent for a planned, cited report in Canvas.
Babbily can use search tools when a question needs current information, source links, or broader research.
## Web search
Use web search for focused questions that need current sources.
Good examples:
* "Search the web for the latest pricing changes from this vendor."
* "Find recent articles about this market and summarize the pattern."
* "Compare these two tools using current public information."
* "Show me photos of the new stadium and summarize the latest construction updates."
Search replies can include citations so you can review the sources.
Search can also add a few pictures above the answer when the subject is something you can see or when you ask to see it. For the first question in a chat, Babbily decides whether pictures help. In follow-up questions, Babbily adds search pictures only when you ask for them, and it leaves out logos, icons, ads, and pictures that do not match your question.
Search replies may include citations, pictures of a visual subject, charts, or other visual formats when they help. Learn how to ask for pictures from web search.
## Deep Research Agent
Use **Deep research** when you need more than a quick lookup. The Deep Research Agent plans a broad question, works through focused sub-questions, gathers sources, and writes a synthesized report into Canvas. The report cites text sources and does not include pictures from web search.
Good examples:
* "Research the competitive landscape for AI meeting assistants."
* "Build a sourced brief on recent changes in enterprise AI procurement."
* "Find the strongest arguments for and against this product direction."
Deep research can take several minutes and may use more API usage budget than a normal reply. Use web search when one focused lookup is enough.
## What happens during a run
The agent turns your request into a concise title, research questions, and a report outline. If the scope is unclear, Babbily may ask you to clarify it first.
The report opens in Canvas before research begins. You can follow the questions being answered, the planned sections, progress, and sources found so far.
Babbily works through the plan with available research tools and selected connected apps. Source availability, blocked pages, and account permissions can affect coverage.
Babbily writes the report into Canvas and returns a short summary in chat. Review the full report and its sources before relying on important conclusions.
## Turn search or research on
Use the tool control beside the message box.
Use **Auto** if you want Babbily to decide whether a tool fits. Use **Manual > Deep research** when you want to run the agent explicitly and it is available for the selected model.
Include the scope, entities, timeframe, sources that matter, and how you want the report structured.
```text theme={null}
Do deep research on the enterprise AI meeting-assistant market. Compare the leading products, pricing signals, security claims, and adoption risks using current sources. Put the full report in Canvas with a source list and open questions.
```
## Approve plugin actions during research
If Deep Research uses a selected plugin and an action needs your approval, Babbily pauses the run and shows an approval card inline in the conversation. The Canvas report keeps the research gathered so far.
Review the plugin, action, content, and destination. Choose **Allow** to run the action and continue the research. Choose **Cancel** to decline the action and end the paused research run; Canvas keeps the sources gathered before the pause.
A Deep Research report can include information from connected apps you chose to use. Review the report before you export it or share the chat.
## When a report ends early
If evidence gathering reaches its time limit, Babbily finishes a report from the available evidence and tells you its coverage is partial. If report writing reaches its time limit, Canvas keeps the written portion and labels the report incomplete. Review what it covers, then narrow the scope and run a follow-up when you need the missing sections.
If the run fails before it can produce a report, try a narrower request, remove an unavailable source or plugin, switch to a supported model, or retry later.
## Read citations carefully
Citations help you check where an answer came from. They are especially important for fast-changing topics, pricing, policies, legal details, medical topics, and investment research.
Search and research can summarize public or authorized connected-app information, but you should verify high-stakes decisions directly with authoritative sources.
Compare web search, deep research, site map, finance research, and other built-in tools.
Follow a step-by-step research workflow with citations.
Use Finance when your research is about markets, companies, earnings, prediction markets, or watchlists.
# File attachments
Source: https://docs.babbily.com/features/file-attachments
Attach files to give Babbily more context for a chat.
Attachments let Babbily use file context in a chat. You can attach supported files when you want help summarizing, analyzing, transforming, or referencing the content.
## What you can attach
File controls appear only when the selected model supports that kind of input. Depending on the model, Babbily can work with:
* Images.
* Documents.
* Audio.
* Video.
Document uploads from the file picker can include common text and document formats such as `.txt`, `.js`, `.ts`, `.tsx`, `.py`, `.html`, `.css`, `.json`, `.csv`, `.md`, `.pdf`, `.doc`, and `.docx`.
Image uploads generally work best as PNG, JPG/JPEG, or WebP when the model needs to read the image. Other broad file categories may appear when the selected model supports them, but not every model can read every format.
If a file type is hidden or rejected, switch to a model with the right file support or convert the file to a common format. **Fast mode** can change the supported inputs; check the attachment controls after turning it on.
## Attach a file
Start a new chat or continue an existing thread.
Use the attachment control near the composer and choose the file.
Ask a specific question about the file.
```text theme={null}
Summarize this document for an executive reader and list the open decisions.
```
Babbily uses the file if the selected model supports it.
When editing an earlier message, its original attachments must still be available and compatible with the selected model. See [editing earlier messages](/features/chat-and-models#edit-an-earlier-message) before replacing a message.
## Tips for better file results
* Choose a model that supports the file category before attaching.
* Ask for the output format you want.
* Mention which part of the file matters most.
* Use one or a few relevant files instead of attaching everything.
* Convert unusual files to a common format if the upload is rejected.
* Start a new thread when the file is no longer relevant.
## Reuse files across a Project
A normal attachment belongs to the chat where you add it. If reference material should be available across several related chats, add it to a Project's **Knowledge** card instead.
Project knowledge supports PDF, DOCX, plain text, Markdown, CSV, and JSON files up to 25 MB each, with up to 100 files per Project.
Add reusable reference material and understand processing and context limits.
Do not upload passwords, API keys, private keys, payment card numbers, or other secrets. Only attach files you are allowed to use in Babbily.
Troubleshoot unsupported files, browser permissions, device behavior, and voice input.
# Finance
Source: https://docs.babbily.com/features/finance
Use Babbily Finance for market overviews, company pages, watchlists, earnings, prediction markets, and source-aware finance chat.
Finance gives you a dedicated place to review markets, open company pages, save a watchlist, and ask follow-up questions with market context attached.
Finance content in Babbily is informational. Verify important details with original sources and do not treat Babbily as financial, investment, tax, or legal advice.
## Open Finance
Open **Finance** from the sidebar, or select **Finance** from the News topic tabs.
Finance includes tabs for:
* **US Markets**.
* **Crypto**.
* **Earnings**.
* **Predictions**.
Availability can vary by market, company, source availability, plan, and account state.
## Review market overviews
Use market overviews when you want a quick read on current market conditions.
Market pages can include:
* A ticker bar with active market rows.
* Sentiment and market status.
* Top assets or popular cryptocurrencies.
* Market summary and source-backed market stories.
* Sector heatmaps and standout charts.
* Movers, fixed income, peers, and related market rows when available.
Select a market row or company to open a company page.
## Search and save a watchlist
Use the Finance search box to look up stocks, crypto, and other supported market symbols.
On company and market rows, use the star button to add or remove a company from your **Watchlist**. If your watchlist is empty, Babbily shows trending companies as suggestions.
## Open company pages
Company pages can include:
* Current price, daily movement, session timing, and recent update time.
* Chart ranges such as intraday and longer historical windows.
* Key stats.
* Notable movement and source-backed stories.
* Price movement timeline.
* Key issues with bullish and bearish views.
* Company profile, analyst consensus, and peers when available.
Company pages also include tabs for **Overview**, **Predictions**, and **Historical Data** when supported.
## Export historical data
Open a company page, then choose **Historical Data**. Use the interval and range controls to change the table. When rows are available, you can download the data as **CSV** or **PDF**.
Historical data may be incomplete or temporarily unavailable for some symbols, intervals, or market states.
## Explore earnings and prediction markets
Use **Earnings** to review upcoming or recent company earnings events. Open an event to see available company context, estimates, reported values, and source details.
Use **Predictions** to browse supported prediction-market cards. You can filter or search prediction markets, open a market detail page, and ask questions about the market context.
## Ask Finance follow-up questions
Finance pages include a composer for source-aware follow-up chat. Ask about the current market overview, a company page, an earnings event, or a prediction market, and Babbily attaches the page context to the new chat.
Examples:
```text theme={null}
Summarize the biggest market movers and list what I should verify.
```
```text theme={null}
Explain the bullish and bearish cases for this company using the available sources.
```
```text theme={null}
Compare this prediction market with recent company and market context.
```
You can also use **Manual > Finance research** from a regular chat when the selected model and your plan support it.
## Choose how Finance research runs
Finance threads keep their page context and tool setting through follow-up questions, page reloads, and **Retry**.
| Tool mode | What happens in a supported Finance thread |
| ---------- | ------------------------------------------------------------------------------------------------------------------ |
| **Auto** | Babbily uses Finance research automatically for Finance follow-ups. |
| **Manual** | **Finance research** starts enabled for the Finance thread. You can remove it when you do not want that tool used. |
| **Off** | Babbily does not call tools and gives a model-only reply. |
If the selected model does not support Finance research, Babbily keeps your model selection and does not expose the unsupported tool. Switch to a supported model when you need source-aware Finance research.
These defaults apply to a chat started from Finance context. They do not turn Finance research on for unrelated chats.
## Research filings, periods, and comparisons
Finance research can help with:
* Live quotes and daily market ranges.
* Historical company performance and fiscal periods.
* Earnings context and company filings.
* Peer and multi-company comparisons.
* Revenue, reported operating results, growth, margins, and other available financial metrics.
For comparisons, name each company clearly and include the period and metrics you need. Babbily accounts for company-specific fiscal calendars and currencies when the available sources support them.
```text theme={null}
Compare the last three reported fiscal years of revenue and operating margin for Microsoft and Alphabet. Cite the filing used for each company and year.
```
Finance research checks whether the available evidence covers the companies, periods, and metrics you requested. When a primary filing can fill a gap, Babbily can use it to complete the answer. Citations stay attached to the answer produced from those sources.
Companies do not all report the same line items. If a requested metric is not reported separately and cannot be derived from primary sources, Babbily should explain the limitation instead of inventing a value. Ask for a related reported figure when one is suggested.
## When data is unavailable
Finance relies on current market and source data. If a panel is temporarily unavailable:
* Refresh the page.
* Try a broader market tab.
* Search for the company by symbol.
* Use a precise company name or symbol if the name could refer to more than one issuer.
* Open cited stories or source links directly.
* Ask Babbily to explain what data is present and what is missing.
* Use **Retry** after a temporary research failure; the Finance context and tool setting remain with the thread.
Browse News, review citations, and open Finance from market and company links.
Compare Finance research with web search, deep research, site map, weather, and other tools.
Learn when to use source-backed research tools and how to review citations.
Understand what can vary by source availability, model, tool, plan, usage, and market state.
# Image and video generation
Source: https://docs.babbily.com/features/image-and-video-generation
Create images and short videos from prompts or supported source media.
Babbily can create images and short videos when your selected model and plan support media generation.
Learn how generated media fits alongside maps, charts, QR codes, citations, and other visual chat replies.
## Create an image
Select a model that supports image generation, or use **Manual** tool mode and choose image generation. Choose a lighter option such as **Nano Banana 2 Lite** when you want faster iteration, and choose a stronger image model when quality or detail matters more.
Include subject, style, composition, lighting, aspect ratio, and any constraints.
```text theme={null}
Create a clean product illustration of a collaborative AI workspace, warm lighting, minimal background, 16:9.
```
Save, download, or continue refining the image in the chat.
## Choose the image model or chat tool
**GPT Image 2.5 Flare** is available in the model picker for image generation and editing. To edit an image, select a model that supports source images, attach the image, and describe the change. Supported source files and controls vary by model.
The built-in **Image generation** chat tool also uses GPT Image 2.5 Flare, but creates a new image from your text prompt. It does not accept an attached image as an editing input. Use a supported image model directly for that workflow.
For the chat tool, you can request square, portrait, or landscape output. Its supported sizes are 1024 × 1024, 1024 × 1536, and 1536 × 1024. Other image models may offer different aspect ratios.
**GPT Image 2** remains selectable as **Sunsetting**. Choose **GPT Image 2.5 Flare** for new work.
## Create a video
Video generation works best with clear motion, subject, and duration instructions.
```text theme={null}
Create a short video of a product dashboard animating from empty state to completed weekly report.
```
If source media is supported for the selected model, you can attach an eligible image or video and ask Babbily to use it. Some media models require a source image or video before you can send. If Babbily asks for source media, attach the required file and try again.
## If your previous media model is missing
**Seedance 2.0** and **FLUX 2 Pro** were removed from the model picker on September 1, 2026. You can choose **Seedance 2.5** for new video work, or **FLUX Kontext Pro** or **FLUX Kontext Max** for image work. Supported inputs and controls can differ, so review the selected model before generating.
**Nano Banana 2** and **Nano Banana 2 Lite** remain available for image generation. The Gemini Flash chat-model changes do not remove these image models.
See [model transitions and alternatives](/features/model-capabilities#sunsetting-and-removed-models) for the full list of removed options.
## Wait for longer media runs
Image and video generation can take longer than short chat replies. If **Chat completion notifications** are on, Babbily can notify you when an image or video finishes while the tab is inactive or in the background.
Turn notifications on or off from **Settings > Account > Notifications**.
## Prompt tips
* Describe what should be in the frame.
* State the style and mood.
* Include aspect ratio or size when the control is available.
* For video, describe motion and pacing.
* Keep brand-sensitive or regulated imagery specific and review outputs carefully.
## Usage impact
Image and video generation usually use more of your API usage budget than short text replies. If you are near your limit, check the usage card before generating media.
Review generated media before using it publicly. Make sure it fits your brand, rights, and compliance needs.
Follow a step-by-step workflow for creating images or videos from prompts and supported source media.
# Library and sharing
Source: https://docs.babbily.com/features/library-and-sharing
Find saved chats, translations, media, Canvas artifacts, and shareable conversations.
The Library helps you find work you have already created in Babbily. It can include saved chats, translations, Canvas artifacts, images, videos, and other media.
## Find saved work
Open **Library** to browse and search saved items in the active workspace. Switch between **Personal** and a team workspace before opening Library when you need to find work saved somewhere else.
You can filter by content type and use item actions to reopen, rename, share, download, or delete supported items.
The content filters are **Threads**, **Images**, **Audio**, **Videos**, **Translations**, and **Files & Artifacts**.
Use the sort control for **Recent activity**, **Oldest first**, **Name A–Z**, or **Name Z–A**. On smaller screens, some controls use icons; open the item's actions to see available options.
For your eligible threads, the actions also include **Add to project**, **Move to project**, and **Remove from project**. See [organizing existing threads](/features/projects#add-an-existing-thread).
Scheduled Task results open from **Scheduled > Recent runs**, rather than the normal Library thread list.
## Find Canvas artifacts
Canvas artifacts are saved pieces of generated work, such as drafts, tables, code, image briefs, and structured outputs. Open an artifact from Library to review it, continue editing, download supported formats, or send it to a connected app when actions are available.
Learn how to edit, version, export, and send Canvas artifacts.
## Share inside a team workspace
When you author a supported chat or saved media item in a team workspace, its share controls can let you choose:
* A specific teammate.
* **Everyone in** the team.
Recipients find the item under **Shared with you** in the matching team workspace's Library. Shared team chats are read-only, so recipients can review the conversation but cannot reply in the original thread.
Team sharing is membership-gated. If someone leaves the team, their access to direct and team-wide shares ends.
Team sharing is different from a public share link. It keeps access inside the team and does not make the item public.
## Share a chat with a public link
When sharing is available for a chat, use the share action on the thread. Babbily creates a public share view that others can open with the share link.
Temporary Chats are not shareable. Use a regular chat if you need to create a share link.
Shared chats are useful for:
* Sending a research summary to a teammate.
* Sharing a generated answer without inviting someone into your account.
* Showing a public version of a conversation.
## What shared viewers can see
People with the link can see the messages plus any attachments, generated media, citations, or other content included in the shared thread. They do not get access to your account, private settings, plugins, billing, or memory controls.
Review a thread before sharing it. Do not share chats that include confidential, personal, or sensitive information.
If Babbily asks you to wait before sharing, let the conversation finish its current action and try again. A shared thread does not support [past-chat lookup](/features/memory#refer-to-past-chats) for subsequent replies.
## Public search visibility
The share dialog lets you make a link public or private. When a thread is public, the dialog notes that it can be found in Google Search Results. When you make a link private, Babbily removes the thread from public search visibility where that control applies.
Search engines can take time to update. If a thread included sensitive information, make it private and contact support if you need help.
## Delete saved items
Deleting an item removes it from your Library. Some deletion actions cannot be undone.
If you want a chat that does not appear in Library, use [Temporary Chat](/features/temporary-chat) before you begin. If you want to find saved translation work, open [Translate](/features/translate) or **Library > Translations**.
# Memory
Source: https://docs.babbily.com/features/memory
Control the context Babbily can remember and reuse across chats.
Memory lets Babbily carry useful context from one chat to the next. You can save preferences, work context, recurring projects, and other details you do not want to repeat. This gives recurring work a consistent starting point without re-entering the same setup in every chat.
See what to save to memory, what to keep in the current chat, and what to avoid saving.
## When to use memory
Use memory for context that should help future chats, such as:
* Your role, team, or company.
* Projects you are actively working on.
* Preferred tone, format, or units.
* Reusable work preferences.
* People, products, or accounts that come up often.
For one-time context, keep it in the current message instead of saving it.
## Memory and workspaces
Memory follows the workspace where a chat was created:
* Your **Personal** workspace uses your personal saved memory and Memory Profile.
* Each team workspace keeps a separate team-workspace memory context.
* Team-workspace memories do not change or expose your personal Memory Profile.
By default, your memory inside a team workspace stays private to you. A team owner can turn **Shared team memory** on from **Settings > Team**. When it is on, useful facts from team-workspace chats can be added to shared context that current members can use in team chats.
Authors and the team owner can delete eligible shared entries. Turning **Shared team memory** off stops new additions; it does not mean existing shared context is copied into anyone's personal Memory Profile.
## Project memory
Each Project can keep a separate memory container for chats in that project. Project memory replaces personal or team-workspace memory for those chats instead of combining the scopes.
Personal Projects start with project memory on. Team Projects start with it off, and the team owner decides whether to enable one shared project memory for current members. Project memories are not copied into your personal Memory Profile.
Learn who can enable project memory, use it, and remove eligible items.
Shared team memory is meant for context every current team member is allowed to use. Keep secrets and information that is not approved for team-wide access out of team chats.
## Add a memory
The fastest way to save context is to ask in chat:
```text theme={null}
Remember that I prefer direct answers with a short summary at the end.
```
You can also add memory from **Settings > Memory** by using the memory instruction field. When available, select **Remember** for a message to request that its stated fact be saved.
With memory enabled, Babbily may also save useful, non-sensitive context from ordinary conversation. Sensitive personal facts need an explicit request and must be relevant when used. Secrets such as passwords and private keys are not appropriate to save.
Check the memory status beneath the reply or in settings. A conversational acknowledgment alone is not confirmation that a memory was saved. A failed or unavailable save may need another attempt. If Babbily says memory is temporarily unavailable, wait a moment and try again. Contact support if it continues.
## Review your Memory Profile
Open **Settings > Memory** to see your personal Memory Profile. Babbily organizes saved context into sections such as work context, personal context, top-of-mind items, recent context, and long-term background.
The profile can change as you add, update, or remove context. Its purpose is to make memory readable so you can understand what Babbily may use.
## Review older saved memories
If **Settings > Memory** shows **Needs review**, those older items are excluded from memory used in replies until you approve them.
* Choose **Approve** to keep an item or **Forget** to remove it from saved memory.
* Use **Approve all** or **Forget all** for the items awaiting review, and read the confirmation before removing them.
* Approval can take time to become available for recall. Check the result message; some items may remain unapproved or need another review.
Review the contents before approving. Bulk approval does not make secrets safe to store.
## Memory citations
When Babbily saves or updates memory from a chat, replies may show a memory citation. A citation gives you a short explanation of what changed.
In **Settings > Memory**, you can expand **Show Me All Memory Citations** to review saved sources and remove eligible items.
## Forget or delete memory
You can remove saved context in two ways:
* Say what to forget in chat or in **Settings > Memory**.
* Delete an eligible citation from the Memory page.
After a confirmed removal, Babbily stops using the item as saved memory. This does not delete the original conversation: the same information may remain in the current thread or eligible past-chat results. Review or delete the original chat separately if you also want to remove that history.
If a source cannot be deleted from the page, contact support for help removing account-level memory.
Do not save passwords, API keys, verification codes, payment cards, or other secrets in memory. Babbily is designed to avoid saving sensitive items, but you should still keep secrets out of chat.
## Refer to past chats
When an earlier discussion matters, ask a specific question in **Auto** mode:
```text theme={null}
What did we decide about the onboarding outline in our earlier chat? Link the conversation if you find it.
```
Past-chat lookup requires memory to be on and a model that supports memory. It looks within your eligible chat history for the current workspace or personal project. It does not search every chat in your account or other people's private conversations.
Past-chat lookup is unavailable in **Manual** or **Off** mode, Temporary Chat, Translate, unattended Scheduled Tasks, team projects, and chats that are publicly or directly shared. A personal project also needs project memory enabled. Some chats started with outside source context may be ineligible.
Search results can be incomplete. An empty result does not prove that a conversation never happened. Open **Library** to find the chat yourself if needed. Historical messages may be outdated; use them as a record of what was discussed, not proof that a fact is still current.
## Turn memory on or off
Open **Settings > Profile** to find the memory preference. When memory is on, Babbily uses your Memory Profile for new replies when the selected model and account state support memory. When memory is off, Babbily does not use saved profile memory to personalize new replies. You can turn it back on when you want Babbily to use your saved context again.
Memory works best for stable preferences. If a detail is temporary, mention it in the current chat instead of saving it.
# Memory examples
Source: https://docs.babbily.com/features/memory-examples
See what to save to memory, what to keep in the current chat, and what to avoid saving.
Memory works best when it captures stable context you want Babbily to reuse across chats. Use this guide to decide what belongs in memory and how to phrase it.
## Good memory examples
| Save this kind of context | Example |
| ------------------------- | --------------------------------------------------------------------------- |
| Communication preference | `Remember that I prefer concise answers with clear next steps.` |
| Role or team context | `Remember that I manage customer operations for a B2B software company.` |
| Recurring project context | `Remember that Project Atlas is our customer onboarding refresh.` |
| Writing style | `Remember that customer emails should sound warm, direct, and low-jargon.` |
| Units or formatting | `Remember that I prefer dates in Month Day, Year format.` |
| Ongoing audience context | `Remember that my weekly updates are written for non-technical executives.` |
## Keep this in the current chat
Some details are useful once but should not become long-term memory.
Examples:
* A one-time deadline.
* A temporary draft direction.
* A single customer's private details.
* A file-specific instruction.
* A preference that only applies to this project phase.
```text theme={null}
For this chat only, write the proposal in a more formal tone.
```
## Do not save sensitive information
Do not ask Babbily to remember:
* Passwords.
* Verification codes.
* API keys.
* Private keys.
* Payment card numbers.
* Government identifiers.
* Raw connected-app credentials.
* Confidential details you are not allowed to use in Babbily.
Memory is designed for useful context, not secrets. Keep sensitive information out of chat and memory.
## Update memory when context changes
When a preference or project changes, say so directly.
```text theme={null}
Update my memory: for customer updates, use a shorter summary first and put details under headings.
```
You can also open **Settings > Memory** to review your Memory Profile and remove eligible memory citations.
## Ask Babbily what it remembers
Use **Settings > Memory** to review your personal Memory Profile, citations, and any items marked **Needs review**. Project memory is reviewed in its project. You can also ask in chat:
```text theme={null}
What do you remember that might affect how you answer my work questions?
```
If something is outdated, ask Babbily to forget it:
```text theme={null}
Forget that Project Atlas is my top priority.
```
Wait for confirmation that a save or removal succeeded. Forgetting a memory does not delete the chat where you originally shared it. See [memory removal and history](/features/memory#forget-or-delete-memory).
## Memory quality checklist
Before saving something, ask:
* Will this still be true in future chats?
* Would this improve many replies, not just one?
* Is it safe to reuse?
* Is it phrased clearly?
* Would I be comfortable seeing it in **Settings > Memory**?
Learn how to review, update, and remove memory.
Review privacy controls for memory, sharing, plugins, and Library items.
# Model capabilities
Source: https://docs.babbily.com/features/model-capabilities
Choose a Babbily model based on speed, reasoning, files, tools, media, context, and API usage budget.
Babbily gives you access to multiple models from one chat surface. The best model depends on what you are trying to do, how much context you need, and which capabilities are available for your plan.
Model availability, names, and capabilities can change over time. Use the model picker as the source of truth for what is available in your account.
## Choose by task
| If you need to... | Look for a model that is strong at... | Helpful setup |
| ------------------------------------ | -------------------------------------------- | -------------------------------------------------------------------------------------- |
| Draft, rewrite, or summarize quickly | Speed and clear writing | Keep tools **Off** unless the answer needs current information. |
| Reason through a complex plan | Deeper reasoning | Give goals, constraints, and decision criteria. |
| Analyze files | File support and enough context | Attach the most relevant files first. |
| Use current sources | Tool support | Use **Auto** or **Manual > Web search**. |
| Produce a broad research brief | Research support | Use **Manual > Deep research**, follow the report in Canvas, and review its citations. |
| Create or edit images | Image generation support | Describe subject, style, composition, lighting, and aspect ratio. |
| Create short videos | Video generation support | Describe motion, pacing, duration, and source media when supported. |
| Compare multiple approaches | Multi-perspective reasoning and synthesis | Ask for options, tradeoffs, and a final recommendation. |
| Work across a larger codebase | Long-context coding and repo-scale reasoning | Attach or paste the most relevant files and define the change boundary. |
| Work in a long thread | Larger context window | Watch the context indicator and start a new thread when the topic changes. |
| Save API usage budget | Lightweight model and fewer tools | Use model-only chat for simple drafts and rewrites. |
## Recently added models
Use the model picker as the source of truth for your account, but these recent additions can help when you are choosing quickly:
As of September 9, 2026, recent additions include **GPT-6 Astra**, **Gemini 3.8 Flash**, and **GPT Image 2.5 Flare**. **GLM 5.3 Flash** now supports **Fast mode**. Available tools and file limits depend on your selected model, plan, and account state.
| Model | Best for | Notes |
| ------------------- | ------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| GPT-6 Astra | Complex reasoning, coding, and extended research | Supports image and document attachments, an approximately 1-million-token context window, **Fast mode**, and reasoning effort through **Max**. It is not a Translate model. |
| GPT Image 2.5 Flare | Everyday image generation and editing | Select it for image work; see [image generation](/features/image-and-video-generation) for the difference between the model and chat tool. |
| Claude Fable 5.1 | Long-running coding, knowledge work, research, and precise instruction following | Choose it for sustained multi-step work. Supports a **Reasoning effort** selector, including **X-High**. |
| GLM 5.3 Flash | Visual coding, tool use, and work across code, documents, and images | Supports **Fast mode** and a **Reasoning** toggle without a separate effort-level selector. |
| Gemini 3.8 Flash | Coding, long-context reasoning, and work with images, documents, and audio | Supports an approximately 1-million-token context window and a **Reasoning** toggle. Also available in **Translate**. |
| Claude Opus 5 | Agentic coding, professional knowledge work, long-horizon reasoning, and vision-enabled workflows | Includes a 1-million-token context window. Turn on **Fast mode** for lower-latency supported replies; it uses more API usage budget. |
| Kimi K3 | Long-horizon coding, agentic workflows, and end-to-end knowledge work | Supports a large context window for tasks that need sustained work across extensive material. |
| GPT-5.6 Sol | Demanding long-horizon reasoning, coding, and complex agentic work | Choose Sol when capability matters more than speed or API usage budget. |
| GPT-5.6 Terra | Everyday reasoning, coding, and multi-step work | Choose Terra when you want a balance of strong capability and API usage budget. |
| GPT-5.6 Luna | Fast drafts, iteration, and lower-cost everyday work | Choose Luna when speed and API usage budget matter most within the GPT-5.6 family. |
| Grok 4.6 | Long-running agentic work, complex research, and coding | Choose it for multi-step investigation and codebase work. |
| Muse Spark 1.2 | Coding, debugging, and understanding large codebases | Supports extended workflows across documents and images. |
## Capability signals
The model picker and composer can show what a model supports. Common capability differences include:
* File types and file size behavior.
* Context window size.
* Tool availability.
* Image generation or editing.
* Video generation or editing.
* Memory personalization availability.
* Speed and reasoning depth.
* Multi-step or agentic task behavior.
* Codebase-scale context handling.
* Fast mode availability.
* Sunsetting status for older model options.
* API usage budget impact.
If a file type, tool, or media action is hidden, the selected model may not support it.
## Sunsetting and removed models
A model that still appears in the picker with a **Sunsetting** label remains selectable during its transition. Choose a newer option for new work. Once an older model is removed from the picker, you can no longer select it there, even if its name still appears in saved chat history.
As of September 9, 2026, these six models remain selectable as **Sunsetting**:
| Sunsetting model | Choose for new work |
| --------------------- | ------------------- |
| Claude Fable 5 | Claude Fable 5.1 |
| GLM 5.2 | GLM 5.3 Flash |
| Gemini 3.6 Flash | Gemini 3.8 Flash |
| Gemini 3.5 Flash Lite | Gemini 3.8 Flash |
| Gemini 3.7 Flash | Gemini 3.8 Flash |
| GPT Image 2 | GPT Image 2.5 Flare |
GLM 5.2 retains **Fast mode** during this transition. Gemini 3.6 Flash, Gemini 3.5 Flash Lite, and Gemini 3.7 Flash remain in the chat picker but are no longer options in **Translate**.
The following older options were removed from the picker on September 1:
| Removed options | Current alternatives |
| ----------------------------------------------------------- | --------------------------------------------------------------- |
| Claude Opus 4.7 and Claude Opus 4.8 | Claude Opus 5 |
| Claude Sonnet 4.6 | Claude Sonnet 5 |
| GPT-5.5 | GPT-5.6 Sol, Terra, or Luna; GPT-5.5 Pro also remains available |
| Grok Build 0.1, Grok 4.3, and Grok 4.5 | Grok 4.6 |
| Gemini 3.1 Flash, Gemini 3.1 Flash Lite, and Gemini 3.1 Pro | Gemini 3.8 Flash |
| Kimi K2.7 Code | Kimi K3; Kimi K2.6 also remains available |
| Muse Spark 1.1 | Muse Spark 1.2 |
| Qwen 3.6 Plus | Qwen 3.8 Max; Qwen 3.7 Max also remains available |
| Seedance 2.0 | Seedance 2.5 |
| FLUX 2 Pro | FLUX Kontext Pro or FLUX Kontext Max for image work |
These are alternatives for new work, not promises of identical output or automatic replacements in every chat. Recheck file support, tools, and API usage budget when switching models.
## Use Fast mode
Fast mode is available only when your selected model supports it. A lightning icon next to a model name means Fast mode is available for that model.
**GPT-6 Astra**, **GLM 5.3 Flash**, **Claude Opus 5**, and **Veo 3.1** support **Fast mode**, as does the sunsetting **GLM 5.2**. For GLM 5.3 Fast, select **GLM 5.3 Flash**, then turn on **Fast mode**; it is not a separate picker choice. The selected model name stays visible. Fast mode can have different file support from the standard mode; check attachment controls before sending.
GLM 5.3 Fast does not directly analyze image, document, audio, or video attachments. Turn **Fast mode** off to use GLM 5.3 Flash for file analysis. If an existing thread contains incompatible attachments, turn Fast mode off or start a new thread.
Claude Fable 5.1 and Gemini 3.8 Flash do not have a separate Fast mode control. “Flash” is part of a model's name, not an indication that it has this toggle.
Open the model picker and choose a model with the Fast mode lightning icon, such as Claude Opus 5.
Choose **Fast mode** using the lightning control near the composer. Babbily shows a usage notice the first time you turn it on.
Keep working with the selected model using its faster response path.
Fast mode can respond faster, but it uses more of your API usage budget. Babbily turns Fast mode off when you change chats or select another model, so check the control before your next prompt.
## Choose reasoning effort
Open the composer menu, choose **Manual**, and turn on **Reasoning**. If the selected model supports effort levels, a **Reasoning effort** selector appears beside the microphone.
| Model | Available controls |
| ---------------- | ---------------------------------------------------------------------------------- |
| GPT-6 Astra | **Low**, **Medium**, **High**, **X-High**, and **Max**, including in Fast mode |
| Claude Fable 5.1 | **Adaptive**, **Low**, **Medium**, **High**, and **X-High** |
| Gemini 3.8 Flash | **Reasoning** on or off; no separate effort-level selector |
| GLM 5.3 Flash | **Reasoning** on or off; no separate effort-level selector, including in Fast mode |
In **Auto**, Babbily chooses reasoning effort for supported work. In **Off**, Babbily does not request optional reasoning. Higher effort can take longer and use more API usage budget. The available levels can change when you switch models.
## Advanced reasoning and coding work
Some models are built for specialized advanced work, such as comparing multiple approaches, reviewing code, working across large files, or planning multi-step engineering tasks. Choose these models when the task benefits from depth, synthesis, or a larger working context.
For better results:
* State the goal, constraints, and expected output.
* Include only the files or excerpts the model needs.
* Ask for a plan before large edits or refactors.
* Ask the model to call out uncertainty, risks, and tradeoffs.
## Switch models during work
You can switch models when the task changes.
Good switching patterns:
* Start with a faster model to brainstorm.
* Switch to a stronger reasoning model for decisions or plans.
* Switch to a file-capable model before attaching documents.
* Switch to a media-capable model for image or video work.
* Use Fast mode when it appears and you want quicker supported replies, then check your API usage budget for heavier work.
* Start a new thread if older context is steering the answer in the wrong direction.
When you switch models, restate the key goal if the thread is long. That helps the new reply focus on the current task.
## Usage considerations
More advanced model work can use more of your API usage budget, especially when it includes tools, long file context, image generation, video generation, or deep research.
For budget-sensitive work:
* Use shorter prompts when the task is simple.
* Avoid unnecessary file attachments.
* Turn tools **Off** for model-only drafting.
* Use search or deep research only when sources matter.
* Check your usage card before running image or video generation.
## Troubleshooting
If a capability is missing:
* Switch to a model that supports it.
* Change tool mode from **Off** to **Auto** or **Manual**.
* Check whether your plan supports the feature.
* Check your remaining API usage budget.
* Try a new thread if the current one has a long or confusing context history.
Learn the basics of choosing models and continuing threads.
Understand why capabilities can vary by model, plan, file, tool, and plugin.
# News
Source: https://docs.babbily.com/features/news
Browse News in Discover, open source-backed articles, set local context, and continue from an article in chat.
News helps you browse current stories in Babbily, review sources, and keep going from an article in chat.
## Browse News
Open **News** from the sidebar. In the app, the page may appear as **Discover**. Babbily shows a feed based on your selected topic and location.
Available categories include:
* Top.
* Tech & Science.
* Business.
* Arts & Culture.
* Finance.
* Sports.
* Entertainment.
The **Finance** tab opens the dedicated [Finance](/features/finance) area instead of a normal news feed.
When you select **Sports**, Babbily may show a live scores ticker above the news feed when current sports data is available.
## Set your location
News can use local context to make the feed more relevant.
Select **News** from the sidebar.
Use the location field near the topic tabs to enter a city or country.
Click the location icon to use your current device or browser location when permissions allow it.
Use the clear control when you want to return to a broader country-level feed.
Your saved profile location can also help Babbily choose relevant News results. You can review it in **Settings > Profile**.
## Use the News side rail
The News side rail can show helpful context beside the feed:
* **Weather** for the selected location.
* **Market Outlook** with market rows that open Finance pages.
* **Trending Companies** with company links, prices, and daily movement when available.
These panels can be temporarily unavailable if current data cannot be reached. Your main News feed may still work while a side rail panel refreshes.
## Open an article
Select a story to open the article view. Babbily can generate a readable article from the available headline, summary, metadata, and cited sources.
Articles can include:
* A generated article body.
* Source citations.
* An image when one is available.
* The source domain and publish timing when available.
* The story category.
Use citations to open the original source material before relying on important details.
## Continue from an article in chat
After an article loads, use the composer at the bottom of the article to ask follow-up questions. Babbily attaches bounded article context to the new chat so you can ask for a summary, timeline, open questions, or source comparison.
Examples:
```text theme={null}
Summarize the key facts and list anything that still needs verification.
```
```text theme={null}
Compare the claims in this article against the cited sources.
```
## Usage impact
Browsing the News feed uses current app data. Generating a full article uses AI work and can have separate rate limits or availability rules from normal chat. If Babbily shows a usage or availability message, follow the in-product guidance.
## Troubleshooting
If News does not show what you expect:
* Update the location field and refresh News.
* Try a broader location, such as a country instead of a city.
* Switch to **Top** if a category has few current stories.
* Open citations when a generated article mentions a source.
* Wait and retry if News, weather, market rows, or trending companies are temporarily unavailable.
News and generated articles are meant to help you understand current stories quickly. Always verify important decisions with original sources.
Use market overviews, company pages, earnings, prediction markets, and watchlists.
Understand what can vary by tool, source, usage state, account state, and current data availability.
# Plugin capabilities
Source: https://docs.babbily.com/features/plugin-capabilities
See what each Babbily Plugin can help you find, reference, draft, or update.
Use this reference when you know which app you want Babbily to use, but you are not sure what to ask for.
Plugin access depends on the permissions you approve in the connected app. Available actions can vary by your permissions, plugin readiness, and what the connected app makes available to your account. Some actions may ask for confirmation before Babbily runs them.
Ask naturally. For example: "Use Gmail to find the latest thread with this customer and summarize the open questions."
## Google Workspace
Use Google Drive to search and reference files you can access.
Good uses include finding files by topic, summarizing document contents, comparing files, and using Drive files as context for a chat.
Use Gmail to find email threads, summarize conversations, and prepare email drafts or other approval-required email actions when available.
Good uses include finding the latest message from a person, summarizing a long thread, drafting a reply, or turning an email into next steps.
Use Google Sheets to work with spreadsheet data.
Good uses include reading a sheet, summarizing rows, updating structured data, formatting ranges, creating formulas, and preparing sheet actions for review.
Use Google Docs to create, search, and update collaborative documents.
Good uses include finding a document, drafting a new doc, editing text, adding tables, updating sections, and turning chat output into document-ready content.
Use Google Slides to create, inspect, and update presentations.
Good uses include generating slide outlines, retrieving slide content, creating a draft deck, updating slide text, or preparing presentation material from a chat.
Use Google Calendar to understand schedule context and plan meetings.
Good uses include checking event context, finding available times, preparing meeting agendas, and summarizing upcoming calendar commitments.
## Microsoft
Use Microsoft Teams to reference chats, channels, and meeting context.
Good uses include summarizing recent team discussions, finding decisions in a channel, and preparing follow-up notes from meeting context.
Use Excel to work with workbooks, worksheets, tables, formulas, charts, filters, and ranges.
Good uses include summarizing workbook data, preparing formulas, updating table values, reviewing charts, and organizing spreadsheet content.
Use Outlook to find email and calendar context.
Good uses include summarizing email threads, finding calendar details, drafting replies, and preparing approval-required email or scheduling actions when available.
## Collaboration and knowledge
Use Notion to search pages, databases, notes, comments, and workspace content.
Good uses include finding project notes, summarizing database items, preparing page updates, and turning a chat into a Notion-ready draft.
Use Slack to reference channels, messages, and threads.
Good uses include summarizing a channel, finding decisions in a thread, drafting a team update, and preparing approved message actions.
## Work management
Use Asana to reference tasks, projects, and planning context.
Good uses include summarizing project status, finding assigned tasks, preparing task updates, and turning chat decisions into work-management actions.
## Business
Use Stripe to reference customer, billing, payment, and subscription context.
Good uses include finding customer billing context, summarizing subscription status, reviewing payment details you are authorized to access, and preparing customer-support notes.
## Developer
Use GitHub to search repositories, issues, pull requests, and code context.
Good uses include summarizing a pull request, finding related issues, explaining a code area, preparing an issue update, and reviewing repository context.
Use Vercel to reference deployments, logs, projects, and hosting context.
Good uses include checking deployment context, summarizing recent project activity, investigating visible hosting signals, and preparing release notes.
Use Sentry to reference issues, traces, release health, and incident context.
Good uses include summarizing recent issues, grouping error patterns, preparing an incident update, and connecting issue context to code or release work.
Use Supabase to reference authorized project context and developer workflow information.
Good uses include understanding project structure, summarizing schema context, preparing developer notes, and pairing Supabase context with code or product questions.
## Approval and permission reminders
* Babbily can only use plugin access you authorize.
* Connected apps may limit what Babbily can see or do based on your role.
* Babbily may ask you to approve write actions, sends, updates, or workflow changes.
* Removing plugin access does not delete content inside the connected app.
Review plugin actions before approving them, especially when they send messages, update records, change workspace content, or affect billing and customer data.
Start from practical examples for email, documents, slides, and app context.
See app-specific prompt examples and caveats for common plugin workflows.
Review plugin privacy controls, approval expectations, and access removal guidance.
# Plugin deep dives
Source: https://docs.babbily.com/features/plugin-deep-dives
See practical examples for using Babbily with email, files, docs, spreadsheets, chat, project, developer, and business Plugins.
Use these examples when you know which app you want Babbily to use but want a more specific prompt pattern.
Babbily can only use plugin access you authorize. Review approval-required actions before they send messages, update records, change workspace content, or affect customer and billing data.
## Google Workspace
Use Gmail when email threads contain the context you need.
Try:
```text theme={null}
Use Gmail to find the latest thread with this customer. Summarize open questions, blockers, and suggested next steps.
```
Good follow-ups include drafting a reply, finding unanswered messages, or turning a thread into a meeting agenda. Review recipients, subject, body, destination, and any requested action before approving.
Use Google Drive to find and reference files you can access.
Try:
```text theme={null}
Search Google Drive for recent planning docs about this launch and summarize the common themes.
```
Good follow-ups include comparing documents, extracting action items, or using a file as context for a new brief.
Use Google Docs to create, search, or update collaborative documents.
Try:
```text theme={null}
Turn this chat into a Google Docs outline for a customer-facing implementation brief.
```
Review the destination and content before approving create or update actions.
Use Google Sheets when the work is structured around rows, columns, formulas, or ranges.
Try:
```text theme={null}
Use Google Sheets to summarize this pipeline sheet by status, owner, and next action.
```
Ask Babbily to propose formulas, formatting, or updates before approving changes.
Use Google Slides to inspect, draft, or update presentations.
Try:
```text theme={null}
Create a slide outline from this research thread with titles, bullets, and speaker notes.
```
Slide actions work best when you specify audience, number of slides, and the level of polish you need.
Use Google Calendar for schedule context and meeting preparation.
Try:
```text theme={null}
Use Google Calendar to summarize my upcoming customer meetings this week and prepare a prep checklist.
```
Calendar access depends on your connected account permissions and calendar visibility.
## Microsoft
Use Outlook for email and calendar context.
Try:
```text theme={null}
Use Outlook to find recent emails from this partner and draft a concise follow-up.
```
Review recipients, subject, body, destination, and any requested action before approving.
Use Excel for workbook, worksheet, table, formula, chart, filter, and range work.
Try:
```text theme={null}
Use Excel to summarize this workbook by region and flag rows that need follow-up.
```
Ask for a proposed change summary before approving updates to shared workbooks.
Use Microsoft Teams to reference chats, channels, and meeting context.
Try:
```text theme={null}
Use Microsoft Teams to summarize decisions from this project channel and list open owners.
```
Connected-app visibility depends on your role and workspace permissions.
## Collaboration and project work
Use Slack to summarize channels, messages, and threads.
Try:
```text theme={null}
Use Slack to find the latest decision about this launch and summarize who owns each next step.
```
Review any message or channel update before approving it.
Use Notion to search pages, databases, notes, comments, and workspace content.
Try:
```text theme={null}
Use Notion to find project notes for this initiative and turn them into a current status summary.
```
Good follow-ups include drafting page updates, summarizing database items, or preparing a Notion-ready brief.
Use Asana to reference tasks, projects, owners, and deadlines.
Try:
```text theme={null}
Use Asana to summarize this project's current status, blocked tasks, and owners.
```
Review task updates before approving changes.
## Developer and operations
Use GitHub to search repositories, issues, pull requests, and code context.
Try:
```text theme={null}
Use GitHub to summarize this pull request, list risks, and draft review questions.
```
GitHub is useful for code understanding and collaboration, but repository permissions still control what Babbily can access.
Use Vercel to reference deployments, projects, logs, and hosting context available to your account.
Try:
```text theme={null}
Use Vercel to summarize the latest deployment status and prepare release notes for the team.
```
Keep requests focused on visible deployment and project signals, not sensitive operational details.
Use Sentry to reference issues, traces, release health, and incident context.
Try:
```text theme={null}
Use Sentry to summarize recent issues for this project and group them by customer impact.
```
Avoid pasting secrets or sensitive event payloads into chat. Use summaries and customer-safe details where possible.
Use Supabase to reference authorized project context and developer workflow information.
Try:
```text theme={null}
Use Supabase to summarize this project's data structure at a high level and identify questions for the engineering team.
```
Keep sensitive credentials, private keys, and confidential data out of chat.
## Customer and business
Use Stripe to reference authorized customer, billing, payment, and subscription context.
Try:
```text theme={null}
Use Stripe to summarize this customer's subscription status and prepare support notes.
```
Treat billing and payment context carefully. Do not paste payment card numbers or sensitive financial identifiers into chat.
## When a plugin answer seems incomplete
Try these checks:
* Confirm the plugin shows connected in **Settings > Plugins**.
* Check whether your connected-app role can access the item.
* Make the app, folder, channel, project, or person more specific.
* Use **Manual** mode and choose the plugin directly.
* Wait a moment after connecting a new app.
* If the plugin was just connected, wait until it finishes preparing tools.
* Revoke and reauthorize access from the connected app if permissions changed.
See the broader plugin catalog and capability summary.
Review plugin access, approvals, and privacy controls.
# Plugins
Source: https://docs.babbily.com/features/plugins
Connect apps so Babbily can use live context and prepare approved actions.
Plugins let Babbily use context from the apps you already work in. You can ask about email, files, calendars, docs, spreadsheets, projects, code, deployments, billing, and more.
See what each plugin can help you find, reference, draft, or update.
Use app-specific examples for email, files, docs, spreadsheets, chat, project, developer, support, and business tools.
Follow a plugin recipe for summarizing email and preparing next steps.
Send supported Canvas work to connected apps when artifact actions are available.
## Available Plugins
| Plugin | What you can use it for |
| --------------- | ------------------------------------------------------------------------------------------------------------------- |
| Google Drive | Search and reference files. |
| Gmail | Find email threads, summarize messages, and prepare email drafts or other approval-required actions when available. |
| Google Sheets | Read, update, and format spreadsheet data. |
| Google Docs | Create, search, and update documents. |
| Google Slides | Create, inspect, and update presentations. |
| Microsoft Teams | Use chats, channels, and meeting context. |
| Excel | Work with spreadsheets, tables, formulas, charts, and ranges. |
| Notion | Search pages, databases, notes, and workspace content. |
| Google Calendar | Use calendar events and scheduling context. |
| Slack | Reference channels, messages, and threads. |
| Asana | Use tasks, projects, and planning context. |
| Supabase | Use authorized Supabase project context. |
| GitHub | Search repositories, issues, pull requests, and code context. |
| Vercel | Use deployments, logs, projects, and hosting context. |
| Sentry | Reference issues, traces, and release health. |
| Stripe | Use customer, billing, payment, and subscription context. |
| Outlook | Find email and calendar context. |
## Connect an app
Go to **Settings > Plugins**.
Select the app you want to connect.
Follow the hosted sign-in and authorization flow for that app. Babbily shows the Plugin as **Connected** when setup finishes.
Ask in plain language, choose a connected app from the composer, mention it with `@`, or type `/` when plugin-specific commands are available.
After authorization, the Plugin may briefly show **Waiting for authorization** while the connection finishes.
## Approve actions before they run
Some plugins can take action, not just read context. When Babbily prepares an action such as an email draft or workspace update, it asks for your approval before running it.
You can choose **Allow** or **Cancel** from the approval card inline in the conversation. If the action includes editable content, review it before approving.
If a Deep Research run requests an approval-required plugin action, the report pauses with the work gathered so far. Allowing the action lets Babbily continue the same report. Canceling declines the action, ends the paused run, and keeps the sources gathered before the pause in Canvas.
Canvas artifacts can also show plugin actions when Babbily can send the artifact to a connected app. Available destinations depend on the artifact type, your plugin permissions, and whether the plugin tools are ready.
Do not approve a plugin action unless you understand what it will do. Approval-required actions ask for your confirmation before they run.
In a chat that uses shared project instructions, Babbily may ask for approval even when a Plugin only reads information. Review the app and requested access before choosing **Allow**.
## Disconnect an app
Use **Settings > Plugins** to disconnect apps where controls are available. If you do not see a disconnect option, revoke access from the connected app's own settings and contact support if you still need help.
Removing plugin access does not delete content inside the connected app.
Fix setup, authorization, preparation, permission, and approval issues.
# Privacy and data
Source: https://docs.babbily.com/features/privacy-and-data
Understand practical ways to control what you share with Babbily.
Babbily works best when you give it useful context. You still control what you type, upload, save to memory, connect, and share.
Babbily does not use your chats to train generative models by default unless you give explicit permission, and Babbily does not sell your personal information. Babbily may use account information for verification, billing, support, safety, product communications, and aggregated usage insights.
Review practical controls for memory, Temporary Chat, plugins, sharing, Library items, and account help.
Understand what you can copy, download, delete, disconnect, or request through support.
Understand safe handling for chats, files, memory, plugins, sharing, teams, and support.
Prepare for privacy, security, billing, SSO, and procurement review.
## What you control
You choose:
* What messages you send.
* What files you attach.
* What memory you save or delete.
* Whether team-workspace context is appropriate for Shared team memory.
* Which apps you connect.
* Which plugin actions you approve.
* Which chats or saved media you share.
* Which saved items you delete.
## Keep sensitive information out of chat
Do not enter or upload:
* Passwords.
* API keys.
* Private keys.
* Verification codes.
* Payment card numbers.
* Government identifiers.
* Confidential data you are not allowed to use in Babbily.
Treat chat, attachments, memory, and connected apps as work surfaces. Only share information you are allowed to process in Babbily.
## Memory controls
Use **Settings > Memory** to review saved context and remove eligible memory citations. You can also tell Babbily what to forget in chat.
Your personal Memory Profile is separate from team-workspace memory. If a team owner enables **Shared team memory**, useful facts from team-workspace chats can become available to current team members. Check the setting under **Settings > Team** and only use team chats for context the whole team is allowed to reuse.
Forgetting saved memory does not delete the original chat. Review [memory controls](/features/memory#forget-or-delete-memory) and remove chat history separately when needed. Older items marked **Needs review** are excluded from saved-memory use until approved.
## Plugin controls
Use **Settings > Plugins** to connect apps and manage connected-app access where controls are available. If you do not see a disconnect option, revoke access from the connected app's own settings and contact support if you still need help.
Some plugin actions require your approval before they run.
## Sharing controls
Team sharing and public share links have different audiences:
* Team-workspace shares give current team members read-only access to supported chats or saved media. You can share with specific members or everyone in the team.
* Public share links can be opened by anyone with the link and may be discoverable in Google Search Results when public search visibility is enabled.
Review the full item before using either option. A public share can include messages, attachments, generated media, citations, or other content in the shared thread.
## Account help
If you need help removing data you cannot manage from the product UI, contact support.
Learn what to include in a support request and what sensitive details to leave out.
# Projects
Source: https://docs.babbily.com/features/projects
Keep related chats, instructions, knowledge files, and memory together in a personal or team workspace.
Projects give ongoing work a dedicated home. Each project can keep its own chats, response instructions, reference files, and memory without mixing that context into unrelated work.
Keep remembered context inside one project instead of your other chats.
Add reference material that chats in the project can use.
Understand what teammates can read, edit, and continue.
Run recurring prompts in your current workspace on a paid plan.
## Create a project
Switch to **Personal** or the team workspace where the work belongs. A project stays in the workspace where you create it.
Choose **Projects** in the sidebar, then select **New project**.
Add a project name and an optional description of the goal or outcome.
Open the project and send a message from its composer. New chats started there are saved to the project automatically.
Use **Search projects** to find a project by its name or description.
## Add an existing thread
Find your saved thread in the sidebar or **Library > Threads** and open its actions menu.
Select **Add to project**, then search for and select a project. Choose **New project** to create one and add the thread in the same flow.
Open the thread in its project. Subsequent replies can use that project's instructions, knowledge, and enabled memory.
For a thread already in a project, use **Move to project** to choose another or **Remove from project** to keep the thread without a project.
You can file only your own eligible saved chats within their existing workspace. You cannot move a Personal chat into a team workspace this way. Temporary Chats and Scheduled Task result chats are not eligible. The destination must be available for new work; wait for active generation or other thread actions to finish first.
Moving a thread keeps its existing messages. It does not transfer previously saved memories between projects or workspaces. The new project context applies to subsequent work; earlier messages may still contain information from the previous project.
Adding a thread to a team project makes it readable by that project's current team members. Review the entire conversation before adding it.
## Add project instructions
Project instructions shape subsequent replies in chats belonging to that project, including threads you add later. Use them for the audience, tone, role, constraints, required sources, or output format that should remain consistent.
Only the first 4,000 characters of project instructions are sent to the model. If your saved instructions are longer, the editor shows how much will be used. Put the most important requirements first.
Keep project-wide rules in instructions and put one-time details in the current chat.
## Add project knowledge
Choose **Add file** in the project's **Knowledge** card to add reference material. Supported project knowledge includes:
* PDF and DOCX documents.
* Plain text and Markdown files.
* CSV and JSON files.
Each file can be up to 25 MB, and a project can contain up to 100 knowledge files. Babbily reads the file after upload and marks it **Ready** when its contents can be used.
Large collections may not fit into every prompt in full. Babbily manages the available context and can use the most relevant material for the current request. The project capacity indicator shows how much of the model's context the ready files may use.
Scanned PDFs without a readable text layer and legacy `.doc` files are not supported as project knowledge. Convert them to a searchable PDF, DOCX, or text format first.
## Use project memory
Project memory is separate from your personal Memory Profile and from memory in other workspaces or projects.
* Personal projects start with project memory on.
* Team projects start with project memory off. The team owner decides whether to turn it on.
* When team project memory is on, chats in that project contribute to one shared project memory that current team members can use there.
* You can remove eligible memories created from your own chats. A teammate's remembered item is labeled separately.
In the project memory card, items marked **Needs review** are not used as saved memory until approved. You can approve or delete eligible items from your own chats; a teammate must manage their own items. Check the status if an approved item is still being prepared for recall.
Turning project memory off stops it from being used for new replies in that project. It does not copy the project's memories into your personal Memory Profile.
Compare personal, team-workspace, shared team, and project memory.
## Personal and team projects
A personal project belongs to your **Personal** workspace. A team project belongs to the selected team workspace.
In an active team project, accepted members can see the project, read its chats, and contribute to shared project content such as instructions and knowledge files. Chats still have individual authors: teammates can read a chat created by another member, but only its author can continue it.
The team owner controls team project memory and project lifecycle actions. If the team plan becomes inactive, the project remains available as read-only until access is restored.
Do not put information in a team project unless every current member of that workspace is allowed to read it.
## Temporary Chat and projects
Temporary Chat does not save chats to a project. If Temporary Chat is on when you open a project, Babbily asks you to turn it off before using the project composer. Turning it off discards the current temporary conversation.
## Archive or delete a project
Available lifecycle controls depend on the workspace and your role.
**Archive project** permanently makes the project read-only. You cannot restore it, but its existing chats remain in your chat history.
**Delete project** permanently removes the project's metadata, knowledge, and memories. Existing chats remain in your chat history, but they are no longer grouped under the deleted project.
Review the confirmation carefully before continuing. Project knowledge and project memory cannot be recovered after deletion.
## Troubleshoot Projects
If a project is missing or cannot be edited:
* Confirm that you are in the workspace where the project was created.
* Check whether the project is archived.
* For a team project, confirm that you are still a member and that the team plan is active.
* Turn off Temporary Chat before starting a project chat.
* Refresh the project if a knowledge file remains **Processing**.
* Convert unsupported or scanned files to a supported searchable format.
Manage workspace membership, sharing, and team-plan behavior.
Compare one-chat attachments with reusable project knowledge.
# Scheduled Tasks
Source: https://docs.babbily.com/features/scheduled-tasks
Run recurring prompts automatically in a personal or team workspace on a paid plan.
Scheduled Tasks let Babbily run a saved prompt on its own and leave the answer in a new chat. Use them for recurring briefings, inbox or calendar summaries, weekly reviews, and topic monitoring.
Scheduled Tasks are available on paid plans. Each run uses the API usage budget of the workspace where the task is saved.
## Create a Scheduled Task
Switch to **Personal** or the team workspace where the task should run. Tasks do not move between workspaces.
Choose **Scheduled** in the sidebar, then select **New task** or start from a template.
Give the task a clear name and include everything Babbily should know when nobody is present to answer a follow-up question.
Select a supported model, schedule, time, and time zone.
Select the connected Plugins the task may read from. Save the task when the summary matches what you expect.
A workspace can contain up to 10 personal Scheduled Tasks or 25 team Scheduled Tasks.
## Choose a schedule
| Frequency | Behavior |
| --------- | --------------------------------------------------------------------------- |
| Manual | Runs only when you choose **Run now**. |
| Hourly | Runs every hour at `:00`, `:15`, `:30`, or `:45`. |
| Daily | Runs every day at the selected local time. |
| Weekdays | Runs Monday through Friday at the selected local time. |
| Weekly | Runs on the selected weekday and local time. |
| Monthly | Runs on the selected day and local time. Shorter months use their last day. |
Babbily stores the task's time zone and shows its next run using that zone. Daylight-saving changes are handled from the selected local schedule.
## Run, pause, and review a task
Choose **Run now** to start an extra run without waiting for the next scheduled time. Babbily prevents the same task from starting a second run while one is already active.
Use the task switch to pause or resume future runs. Expand **Recent runs** to review queued, running, completed, skipped, failed, or canceled runs and open the chat created by a completed run.
Scheduled run chats do not appear in the normal Library or recent-chat list. Open them from the task's latest result or **Recent runs**. Deleting a task removes that run history from **Scheduled**, so open or save anything you need before deleting it.
If a task's selected model is retired, Babbily can use an available replacement and labels the run accordingly. Edit the task to choose a replacement yourself.
## Use Plugins safely
Scheduled runs are unattended. A selected Plugin can read authorized context, but it cannot perform a write action that would normally pause for your approval.
Scheduled Tasks cannot run Deep Research, artifact actions, or past-chat lookup. Include the context the task needs in its saved prompt. If the task needs to send a message, update a record, change a document, or perform another approval-required action, use an attended chat instead.
A prompt can still be saved when no suitable Plugin is selected. Review any capability warning before saving, especially for calendar, email, or workspace requests.
## Reconnect older Plugin selections
Tasks created with a retired Smithery connection cannot run until you make an explicit current selection.
Open **Settings > Plugins** and connect the current Plugin through the Composio authorization flow.
Open the task showing **Plugin reselection required**.
Choose the connected Plugins it may use now—even if the correct choice is none—and save the task.
Former credentials were not transferred from Smithery to Composio.
## Team Scheduled Tasks
Accepted team members can view, edit, run, pause, and delete tasks in their team workspace. Editing a task changes its **Runs as** identity to the editor, so future runs use that member's connected Plugins.
Everyone in the team can read a run's result. Only the member shown in **Runs as** can continue that generated chat.
If the team plan becomes inactive, members can still pause or delete retained tasks, but new runs do not proceed until plan access is restored.
## Usage and failed runs
Every run checks the workspace plan and available API usage budget before generating. A run can be skipped or fail when:
* The paid plan is no longer active or cannot be verified.
* The workspace API usage budget is exhausted.
* The selected Plugin needs reconnection or reselection.
* The task is already running.
* The selected model is unavailable and no replacement can be used.
Open **Recent runs** for the result and try the suggested action before running the task again.
Connect apps and choose the context a task may read.
Fix authorization, reconnection, and Plugin reselection issues.
Understand how each run uses the selected workspace's budget.
Understand workspace membership and team-plan access.
# Security and procurement
Source: https://docs.babbily.com/features/security-and-procurement
Prepare for a Babbily security, privacy, SSO, billing, and procurement review.
Use this page when your organization needs to review Babbily before buying, expanding, or approving wider usage.
This page summarizes customer-facing product expectations. It is not a legal agreement, security certification, or replacement for Babbily's current Terms, Privacy Policy, or approved procurement materials.
## What to review first
Understand practical ways to control what you share with Babbily.
Review safe handling for chats, files, memory, plugins, sharing, teams, and support.
Review team, SSO, support, privacy, and procurement questions.
Understand what you can copy, download, delete, or request through support.
## Common procurement questions
Babbily's privacy settings state that Babbily does not use your chats to train generative models by default unless you give explicit permission.
Team plan cards may list SSO integration. Availability and setup can depend on the plan and account. Contact support if SSO is a procurement requirement.
Plugin access depends on the account that authorizes the connected app and the permissions available in that app. Workspace admins should also manage app access from the connected app's own admin settings.
Personal and team work appear in distinct workspaces. Chats, Library items, and memory stay with the workspace where they were created. Team sharing is membership-gated, while public share links use separate visibility controls.
No. A personal Memory Profile remains separate. A team owner can enable Shared team memory for team-workspace context, so organizations should define approved team-wide data before turning it on.
Users can delete supported saved items, manage memory, disconnect apps where controls are available, and contact support for account-level deletion or data they cannot remove themselves.
Include the plan you are considering, expected number of users, billing owner email, SSO needs, privacy questions, support expectations, and required timeline.
## Information to collect before contacting support
| Area | Helpful details |
| ------------ | ------------------------------------------------------------------------------------------------------------------------------------------------ |
| Organization | Company name, billing contact, expected user count, and rollout timeline. |
| Access | Whether you need SSO, team ownership, user management, workspace guidance, or procurement approval. |
| Data | Approved use cases, restricted data types, Shared team memory policy, team-sharing expectations, plugin expectations, and deletion requirements. |
| Billing | Preferred plan, team size, purchase timing, and invoice or subscription needs. |
| Support | Required support level, onboarding needs, and internal launch date. |
## Sensitive information boundaries
Do not include passwords, verification codes, API keys, private keys, payment card numbers, government identifiers, or raw connected-app credentials in procurement emails or support tickets. For other confidential files, share them only when support requests them through an approved channel and your organization permits it.
If your organization requires specific legal documents, security questionnaires, or vendor review forms, contact support with the request and the review deadline.
# Team admin guide
Source: https://docs.babbily.com/features/team-admin-guide
Manage team workspaces, identity, invitations, Shared team memory, billing, and API usage budget.
Use this guide if you own or help manage a Babbily team workspace.
Team controls can vary by plan, role, billing state, and account state. Owner-only controls may be hidden for team members.
## Team roles
| Role | What they can usually do |
| ------------ | ------------------------------------------------------------------------------------------ |
| Team owner | Manage the workspace, invitations, member removal, plan changes, billing, and team usage. |
| Team manager | May appear in some team role views where available. Controls can vary by account and plan. |
| Team member | Use the team plan, view team information, and work in Babbily with their own account. |
Team members may see plan and usage information, but subscription changes and owner-level billing actions belong to the team owner.
## Work across multiple workspaces
Every account has a **Personal** workspace. Accepted team memberships add team workspaces to the account menu, and the sidebar pill shows which workspace is active.
Chats, Library items, and memory stay with the workspace where they were created. Teach members to check the workspace pill before starting work because existing chats cannot be moved by switching workspaces.
## Create a team
Choose **Team** or **Teams Ultra** from the plan flow.
Add the team or company name customers should recognize in settings.
Invite enough people to meet the setup requirement. Team setup requires at least three people total, including the owner.
Finish the billing step so the workspace can become active.
Switch to the team workspace, then open **Settings > Team** to confirm the name, logo, members, plan, and Shared team memory setting.
If you already have an individual plan, it can remain active for your Personal workspace. Team billing is managed separately.
## Invite members
Owners can invite one or more teammates by email from **Settings > Team**.
After an invitation is sent:
* Pending invitations can be reviewed from team settings.
* Invitees join after accepting the invitation.
* Accepted members receive access to the team plan and team API usage budget.
* Pending invitations do not receive team access or usage until accepted, but may affect billed seats until accepted, declined, or canceled.
## Manage members and invitations
Use **Settings > Team** to review accepted members and pending invitations. Depending on your role and the invitation state, you may be able to resend, cancel, or remove access.
Removing a member stops team plan access and membership-gated access to team shares. It does not delete content inside the member's Personal workspace or connected apps.
## Manage team sharing
Team-workspace authors can share supported chats and saved media with specific members or everyone in the team. Recipients see these items under **Shared with you** in Library, and shared chats are read-only.
Membership is checked when the recipient opens the share. Removing a member therefore ends access to membership-gated team shares without changing public share links.
## Govern Shared team memory
The team owner controls **Shared team memory** from **Settings > Team**. All members can see whether it is **Enabled** or **Disabled**.
Before enabling it:
* Decide what team context is appropriate to reuse.
* Tell members that useful facts from team-workspace chats can become available to the whole team.
* Keep confidential information that is not approved for team-wide use out of team chats.
* Establish who should review and delete outdated shared entries.
When enabled, authors and the owner can delete eligible shared entries. Turning it off stops new additions but does not share or change anyone's personal Memory Profile.
## Billing and plan changes
Team owners manage billing from **Settings > Billing**. Switch to the team workspace you own before changing that team's plan. Team members should ask the owner to make plan changes, update billing, or troubleshoot missing subscription controls.
If an individual plan and team plan coexist, manage each in its own plan context. Joining or creating a team does not automatically cancel the individual subscription.
If a recent billing change is not visible yet, refresh settings and check again after a short wait.
If a team plan becomes inactive, Babbily marks the workspace **Plan inactive — read-only**. Restore the team subscription before expecting members to create new work there.
## Team API usage budget
Team seats are billed separately from usage availability. Accepted team members receive the team plan's included API usage budget and follow the team plan reset timing.
If a team workspace runs out of included usage:
* Team members should contact the team owner.
* Team owners can review usage from billing and usage surfaces.
* Team owners can upgrade when the plan supports it.
## Troubleshooting team controls
If a team control is missing:
* Confirm whether you are the team owner.
* Confirm you are in the intended team workspace.
* Confirm the invitation was accepted.
* Refresh **Settings > Team** and **Settings > Billing**.
* Ask the team owner to manage billing or subscription changes.
* Contact support if the expected control still does not appear.
Learn the basic team workspace flow.
Understand plan controls and billing visibility.
Review SSO, analytics, priority support, privacy, and procurement questions.
See how usage works for individuals and teams.
Review members, pending invitations, shared links, plugin access, billing, and usage when access changes.
Learn what to include when asking for help.
# Team offboarding and access review
Source: https://docs.babbily.com/features/team-offboarding-and-access-review
Review members, team shares, Shared team memory, public links, plugins, billing, and usage when team access changes.
Use this checklist when someone joins, leaves, changes roles, or no longer needs team access.
Available controls can vary by plan, role, billing state, and account state. Team owners manage most team access and billing actions.
## Offboard a team member
Go to **Settings > Team**.
Make sure you are the team owner or have the controls needed to manage membership.
Find the member whose team access should change.
Use the available member controls. Removed members no longer receive team plan access.
Check **Settings > Billing** and the usage card after the membership change has had time to refresh.
Removing a member ends team plan access and membership-gated access to team shares. It does not delete content inside that person's Personal workspace or connected apps.
## Review pending invitations
Pending invitations do not receive team access until accepted, but they are still worth reviewing.
Check for:
* Invitations sent to the wrong email.
* Invitations that are no longer needed.
* Duplicate invites.
* Teammates who need a resend.
* People who should be invited with a different email.
Cancel or decline invitations that are no longer needed, since pending invitations may affect billed seats before the invitee accepts.
## Review team sharing and memory
Direct and team-wide shares are checked against current membership. When someone is removed, they can no longer open those team shares.
The team owner should still review:
* Chats and saved media shared with **Everyone in** the team.
* Direct shares that are no longer needed.
* Shared team memory contributed during the member's time on the team.
* Whether **Shared team memory** should remain enabled.
Removing a member does not automatically delete shared team memory. The team owner can delete eligible shared entries during the offboarding review.
## Review public share links
Public share links are separate from membership-gated team sharing. When team access changes, ask the owner or relevant teammate to review public share links.
Good checks:
* Does the shared thread still need to be public?
* Does it include customer, billing, plugin, or confidential context?
* Should the thread be replaced with a cleaned-up summary?
* Does the audience still need access?
Shared links do not grant access to account settings, billing, plugins, or memory controls, but anyone with the public link can view the messages plus any attachments, generated media, citations, or other content included in the shared thread.
Public shared threads may also be discoverable in Google Search Results when public search visibility is enabled. Make links private when the thread no longer needs to be public.
## Review plugin access
Plugin access belongs to the account that authorized it. A team membership change does not automatically delete content inside connected apps.
Ask affected users to:
* Review connected apps in **Settings > Plugins**.
* Revoke access from the connected app if needed.
* Remove app permissions from the app's own admin settings when appropriate.
* Contact support if access cannot be managed from the UI.
## Review billing and usage
Team owners should check:
* Whether the member is accepted or pending.
* Whether billing controls show the expected team state.
* Whether API usage budget timing still matches the team plan.
* Whether a plan upgrade or downgrade is needed.
Billing and usage surfaces can take a short time to refresh after membership or plan changes.
## Access review cadence
For healthy team administration, review these regularly:
* Accepted members.
* Pending invitations.
* Billing owner and plan.
* Team-wide and direct shares.
* Shared team memory settings and eligible entries.
* Public shared chats.
* Plugin authorization status.
* Team usage and low-budget warnings.
Manage team setup, invitations, billing visibility, and team usage.
Understand safe handling for chats, memory, plugins, sharing, and support.
# Teams
Source: https://docs.babbily.com/features/teams
Switch team workspaces, collaborate with members, manage Shared team memory, and understand team plans.
Teams give you a dedicated Babbily workspace for team work, membership, and billing. Your **Personal** workspace stays separate, and you can switch among the team workspaces you belong to.
Manage team setup, identity, invitations, Shared team memory, billing visibility, and API usage budget.
Review membership-gated shares, public links, memory, plugins, billing, and usage when access changes.
## Switch workspaces
Your account menu lists **Personal** and the team workspaces available to you. The workspace pill in the sidebar shows the active workspace.
Select the workspace pill in the sidebar or open your account menu.
Select **Personal** or the team name. Babbily opens a clean chat surface in that workspace.
Check the workspace pill before starting a chat or opening Library.
The active workspace scopes your chats, Library items, memory context, and team-sharing options. A chat stays in the workspace where it was created; switching workspaces does not move it.
## Share with teammates
In a team workspace, the author of a supported chat or saved media item can share it with:
* A specific teammate.
* Everyone in the team.
Recipients find team shares under **Shared with you** in Library. Shared chats are read-only, and membership-gated access ends when someone leaves the team.
Team sharing does not make the item public. Public share links use separate visibility controls.
## Shared team memory
**Shared team memory** lets useful facts from team-workspace chats become reusable context for current team members.
* The team owner turns it on or off from **Settings > Team**.
* Every member can see whether it is **Enabled** or **Disabled**.
* When it is on, useful facts from team chats can be added to the shared team context.
* Authors and the team owner can delete eligible shared entries.
* Turning it off stops new additions; it does not mean your personal Memory Profile is shared.
When Shared team memory is off, your team-workspace memory stays private to you. Personal workspace memory always remains separate from team memory.
Only use Shared team memory for information every current team member is allowed to use. Keep passwords, private keys, payment details, and other secrets out of chat and memory.
## Projects and Scheduled Tasks
Team Projects keep related chats, instructions, knowledge files, and optional project memory inside the team workspace. Accepted members can read project chats and contribute to shared project content, but only a chat's author can continue it. The team owner controls team project memory and lifecycle actions.
On a paid team plan, accepted members can view, edit, run, pause, and delete team Scheduled Tasks. Editing a task changes its **Runs as** identity to the editor, so future runs use that member's connected Plugins. Everyone can read the result, but only the **Runs as** member can continue its generated chat.
Organize collaborative work and understand project access.
Automate recurring team prompts and review run ownership.
## What team owners can do
Team owners can:
* Create a team workspace.
* Add or change the team logo and name.
* Invite members.
* Remove members.
* Turn Shared team memory on or off.
* Manage team billing.
* Review team plan status and usage.
Available controls can depend on your plan and account role.
Some workspaces may also show manager roles where available. Owner-only actions remain limited to owners when the control is owner-gated.
## Invite members
Go to **Settings > Team**.
Enter the teammate's email address and send the invitation.
The teammate receives an invitation and joins after accepting it.
Return to **Settings > Team** to see accepted members and pending invitations.
## Team usage
Team plans use seat-based billing with a three-person minimum, including the owner. Accepted team members receive the team plan's included API usage budget and follow the team plan reset timing.
Pending invitations do not receive team access or usage until accepted, but they may affect the billed seat count until accepted, declined, or canceled.
## Team billing
Team billing is managed by the team owner. If you are a team member, billing controls may be hidden or limited. An individual plan can remain active alongside team membership; joining a team does not automatically cancel that separate subscription.
If you own a team, switch to that team workspace before changing its plan. If the team plan becomes inactive, the workspace is labeled **Plan inactive — read-only** until the owner restores access.
Some team billing changes may take time to appear across every billing and usage surface. Refresh settings if a recent change is not visible yet.
Review team setup, SSO, analytics, support, privacy, and procurement questions.
# Temporary Chat
Source: https://docs.babbily.com/features/temporary-chat
Use Temporary Chat when you do not want a conversation saved in your Library.
Temporary Chat is for conversations you do not want saved in normal chat history or Library.
When you start Temporary Chat, Babbily states that the chat will not appear in history, will not use or update Babbily memory, and will not be used to train Babbily's models.
## When to use Temporary Chat
Use Temporary Chat for:
* One-off brainstorming.
* Drafts you do not need later.
* Quick questions.
* Sensitive context you do not want saved in your normal history.
Temporary Chat helps keep a conversation out of saved chat surfaces, but you should still avoid entering passwords, keys, payment details, or other secrets.
## Start Temporary Chat
Use the Temporary Chat control in the chat interface.
Start the temporary session.
You can still use supported models, tools, attachments, image generation, and video generation.
Starting a new chat or leaving the temporary thread exits Temporary Chat and removes it from saved chat surfaces.
## What changes in Temporary Chat
* The thread does not appear in Recents or Library.
* The chat does not use or update Babbily memory, including past-chat lookup.
* The thread is not meant for sharing.
* Active generation may be stopped when you exit.
* The session can restore in the same browser tab while it is active.
* Exiting permanently deletes the current temporary conversation.
## Context limits
Temporary Chats do not use automatic compaction. If a Temporary Chat fills the context window, Babbily may ask you to start a saved thread to keep going.
If you realize you want to keep the work, copy the useful result into a normal chat or document before exiting Temporary Chat.
# Tools and skills
Source: https://docs.babbily.com/features/tools-and-skills
Control when Babbily can use search, research, media generation, connected apps, and reusable skills.
Tools let Babbily do things a model cannot do by itself, such as search the web, run deeper research, make QR codes, generate media, or use a connected app. Skills are reusable instructions that help Babbily follow a repeatable workflow.
## Tool modes
Use the composer menu to choose how much freedom Babbily has for the next reply.
| Mode | What it does | Best for |
| ---------- | --------------------------------------------------------- | -------------------------------------------------------- |
| **Auto** | Babbily decides which available tools fit your message. | Most chats, research, and broad questions. |
| **Manual** | You choose the tool, skill, or plugin Babbily should use. | Forcing search, research, media, or a specific workflow. |
| **Off** | Babbily answers without tools. | Fast model-only replies, brainstorming, and drafts. |
Some tools appear only when the active model and your plan support them.
## Built-in tools
| Tool | What you can use it for |
| --------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------- |
| Web search | Find current information, cite sources, and show pictures of a visual subject when you ask. |
| Deep research | Plan a broad question, research it across many sources, and write a cited report into Canvas. |
| Site map | Review and summarize a public website or domain. |
| [Finance research](/features/finance#ask-finance-follow-up-questions) | Research companies, markets, filings, periods, and financial topics with source-aware Finance context. |
| Weather | Get current conditions and forecasts. |
| Stats | Show supported compact public-data stats; for your own numbers, ask for a stat-style visual reply or use a spreadsheet or file workflow. |
| QR code | Create a QR code for a link or text. |
| Image generation | Create new images from a text prompt with GPT Image 2.5 Flare. |
| Babbily Docs | Look up Babbily help-center instructions and troubleshooting. |
| Video generation | Create short videos from a prompt or supported source media. |
See when to use each built-in tool, example prompts, and what to try when a tool is missing.
## Use a tool manually
Click the tool control beside the message box.
Select **Manual** so you can pick the tool yourself.
Choose the tool you want, such as **Web search**, **Deep research**, **Finance research**, or **Image generation**.
Babbily uses the selected tool for the reply when the model supports it.
**Deep research** can take several minutes and writes the full report into Canvas. **Finance research** follows the separate Auto, Manual, and Off behavior described in the [Finance guide](/features/finance#choose-how-finance-research-runs).
Past-chat lookup is available only in **Auto** mode when memory and the current chat support it. See [memory and earlier conversations](/features/memory#refer-to-past-chats).
## Skills
Skills are reusable instructions for common workflows. They can help Babbily format work consistently, choose helpful tools, or coordinate with connected apps when you ask.
You can manage skills in **Settings > Skills**:
* Turn Babbily skills on or off.
* Search verified skills.
* Add your own markdown skill.
* Trigger a skill with a slash command in chat.
Built-in skills can include workflows such as Research Brief, Meeting Prep, Writing Polish, Data Analysis, and Code Review. The skill catalog can change over time, so use **Settings > Skills** as the source of truth for what is available in your account.
Markdown skills support a name, description, slug, and markdown instructions. Keep markdown skills focused and avoid adding secrets or credentials.
A skill can guide Babbily, but connected-app actions still follow plugin permission rules. If an action requires approval, Babbily asks before running it.
Ask Babbily for maps, charts, stat cards, weather, QR codes, citations, and media.
Start from common workflows and adjust the prompt for your work.
Use market overviews, company pages, watchlists, earnings, prediction markets, and Finance research.
# Translate
Source: https://docs.babbily.com/features/translate
Translate text, choose a style, use voice input, and continue a translation in chat.
Translate gives you a focused place to translate text without setting up a full chat first.
## Translate text
Select **Translate** from the sidebar.
Click the model name to choose an available translation model. Translate has its own model list, which can differ from the chat picker.
Choose the source language and target language. You can also use auto-detect when it is available.
Type or paste the text you want translated.
Babbily translates after you stop typing. You can copy the result or continue in chat.
## Available translation models
As of September 9, 2026, you can choose **GPT-5.6 Sol**, **Claude Haiku 4.5**, or **Gemini 3.8 Flash** in Translate. Use the Translate model picker as the source of truth for your account.
Gemini 3.8 Flash replaces Gemini 3.7 Flash in Translate. Gemini 3.7 Flash, Gemini 3.6 Flash, and Gemini 3.5 Flash Lite remain selectable as **Sunsetting** in chat, but not in Translate. Check the selected model before translating after a catalog update.
## Choose a translation style
Use style chips when you want the translation to sound a certain way:
* More fluent.
* Business formal.
* Simple enough to explain to a child.
* Academic.
Select the same style again to remove it.
## Use voice input
Use the microphone button in the source text panel to dictate text. Voice input depends on browser, device, microphone permission, and account state.
If voice input is unavailable, type or paste your text instead.
## Continue in chat
After a translation appears, use **Continue in chat** to open a new chat with the original text and translated text included as context.
This is useful when you want to:
* Ask why a phrase was translated a certain way.
* Rewrite the translation for a different audience.
* Turn the translation into an email, post, or document.
## Translation history
Signed-in customers can find saved translations in **Library > Translations**.
Translation uses your API usage budget. For long documents, translate one section at a time so you can review tone and formatting as you go.
# Trust, safety, and data boundaries
Source: https://docs.babbily.com/features/trust-safety-and-data-boundaries
Understand safe ways to use Babbily with chats, files, memory, plugins, sharing, teams, and support.
Babbily is a work surface for AI-assisted tasks. You control what you type, upload, remember, connect, approve, share, and delete where controls are available.
Babbily does not use your chats to train generative models by default unless you give explicit permission, and Babbily does not sell your personal information. Babbily may use account information for verification, billing, support, safety, product communications, and aggregated usage insights.
## What Babbily can help with
Babbily can help you:
* Draft and rewrite content.
* Summarize chats, files, and connected-app context.
* Search public sources and cite them.
* Research broad topics.
* Generate images and short videos when supported.
* Use memory to personalize future replies.
* Prepare plugin actions for review and approval.
## What you should not put in Babbily
Do not enter or upload:
* Passwords.
* Verification codes.
* Private keys.
* API keys.
* Payment card numbers.
* Government identifiers.
* Raw connected-app credentials.
* Confidential data you are not allowed to process in Babbily.
If you are unsure whether something is allowed in Babbily, leave it out or ask your administrator.
## Plugin boundaries
Plugins use access you authorize in the connected app. Connected apps may limit what Babbily can see or do based on your role, app permissions, and workspace settings.
Some plugin actions require approval before they run. Review actions carefully when they:
* Send messages or emails.
* Update documents, tasks, tickets, or records.
* Change workspace content.
* Affect customer, billing, payment, or support data.
* Trigger automations or app workflows.
Removing plugin access does not delete content inside the connected app.
If you use a plugin during Deep Research, the resulting Canvas report can contain information from that app even when private connected-app links are not shown as public citations. Review the report before exporting it or sharing the chat.
## Memory boundaries
Memory is for stable context that should help future replies. It is not for secrets.
Use **Settings > Memory** to review your Memory Profile and remove eligible citations. You can also ask Babbily to forget something in chat.
Personal workspace memory remains separate from team-workspace memory. If the team owner enables **Shared team memory**, useful facts from team chats can become reusable context for current members. Only put context in a team chat when the whole team is allowed to reuse it.
Forgetting saved memory does not delete the original chat. Review [memory controls](/features/memory#forget-or-delete-memory) and remove chat history separately when needed. Older items marked **Needs review** are excluded from saved-memory use until approved.
## Sharing boundaries
Team sharing gives current teammates read-only access to supported team-workspace chats or saved media. You can share with specific members or everyone in the team, and access ends when a recipient leaves the team.
Public share links are separate. Anyone with a public share link can view the messages plus any attachments, generated media, citations, or other content included in the shared thread. Shared links do not give viewers access to your account settings, billing, plugins, or memory controls.
Public shared threads may be discoverable in Google Search Results when public search visibility is enabled. Make links private when public access is no longer appropriate.
## Team boundaries
Team owners manage team membership, invitations, Shared team memory, billing, and owner-only controls. Team membership changes can affect plan access, team-share access, and API usage budget, but they do not delete content inside a member's Personal workspace or connected apps.
## Support boundaries
When contacting support, include enough context to investigate the issue, but hide sensitive details in screenshots and do not send passwords, keys, verification codes, payment card numbers, or raw credentials.
Review practical controls for memory, plugins, sharing, Library items, and account help.
See what to include and what not to send when asking for help.
# API usage budget
Source: https://docs.babbily.com/features/usage-budget
Understand how your monthly Babbily usage budget works, what counts, and when it resets.
Every Babbily plan includes a monthly API usage budget. This budget covers supported AI work Babbily runs for you, such as chat replies, model work, media generation, translation, and some tool work.
## Plan budgets
| Plan | Included usage |
| ----------- | ---------------------------: |
| Free | Limited |
| Personal | \$9.99/month |
| Premium | \$19.99/month |
| Ultra | \$149.99/month |
| Team | \$99.99 per team seat/month |
| Teams Ultra | \$149.99 per team seat/month |
## What counts toward usage
Usage can include:
* Chat replies.
* Supported tool calls such as web search or deep research.
* Image generation.
* Video generation.
* Translation.
* File and media work that requires a model call.
Some features, including generated News articles, Finance data, and certain research tools, may use separate limits or availability rules. Check the usage card and in-product messages for the current state.
## Where to see usage
The usage card shows your current usage and remaining budget. You can also open **Settings > Billing** to review your plan and billing details.
Babbily may refresh the usage card after model or media work finishes. If the number looks slightly delayed, wait a moment and refresh.
## Usage and workspaces
Switching workspaces changes where new work is saved. It does not reset your API usage budget, transfer earlier usage, or move an existing chat. A follow-up remains associated with the workspace where that chat was created.
Check the usage card before starting heavier work, especially after joining a team or changing a plan.
Plan, billing, and membership changes can take a short time to appear. Use the visible usage card as the source of truth for the usage currently available to you.
## When usage resets
* Individual plans reset on the usage period shown in your account.
* Paid plans usually align with the billing period.
* Accepted team members share the timing shown for the team plan.
Unused budget does not roll over.
## Low-budget and exhausted states
Babbily warns you when usage is getting high. If your budget is exhausted, new AI work may be blocked until your plan resets or you upgrade.
Existing chats and saved content remain readable.
Image and video generation usually use more budget than short text replies. If you are near your limit, use model-only chat for lightweight tasks.
# Visual replies in chat
Source: https://docs.babbily.com/features/visual-replies-in-chat
Ask Babbily for maps, charts, stat cards, weather, QR codes, code diffs, citations, search pictures, Canvas artifacts, images, and videos.
Babbily can show more than plain text when a visual answer helps. Depending on the model, tool mode, plan, and your request, replies can include maps, charts, stat cards, weather cards, QR codes, code blocks, code diffs, citations, pictures from web search, Canvas artifacts, generated images, and videos.
Be direct about the format you want. For example: "Show this as a bar chart," "Create a QR code," or "Map these locations."
## What Babbily can show
| Visual reply | Best for | Example prompt |
| --------------- | ----------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------- |
| Map | Specific places, routes, or location lists. | `Map these public city offices: Denver, Boulder, Fort Collins.` |
| Chart | Labeled numeric data, comparisons, trends, and breakdowns. | `Make a bar chart from: Q1 120, Q2 175, Q3 160.` |
| Stat card | A small set of clear metrics or KPIs. | `Show these numbers as compact stats cards.` |
| Weather card | Current weather and short forecasts. | `Show the 5-day forecast for Denver in Fahrenheit.` |
| QR code | Public links or non-sensitive text. | `Create a QR code for https://babbily.com.` |
| Code diff | Clear before-and-after code changes. | `Show the proposed change as a code diff.` |
| Citations | Search, research, and source-backed answers. | `Search recent public sources and cite them.` |
| Search pictures | Seeing a place, landmark, product, artwork, animal, vehicle, or dish that a search-backed answer describes. | `Show me photos of the Sagrada Familia interior.` |
| Canvas artifact | Drafts, Deep Research reports, tables, code, image briefs, and structured outputs you want to edit. | `Create this as a Canvas artifact I can revise.` |
| Image or video | Generated media from prompts or supported source files. | `Create a 16:9 product illustration.` |
## Ask for a map
Maps work best when you give Babbily clear locations.
Good prompts:
* `Map these office locations and label each one.`
* `Show a map of these addresses and summarize the travel pattern.`
* `Find public locations for these venues and map them.`
If Babbily returns text instead of a map, make the locations more specific or ask again with "show this as a map."
## Turn data into charts or stat cards
Charts need clean labels and numbers. Stat cards work well for a small group of headline metrics.
Paste a small table, attach a file, or describe the values clearly.
Ask for a bar chart, line chart, pie chart, table, or stat cards.
Tell Babbily what matters: trend, ranking, percent change, totals, or outliers.
```text theme={null}
Create a line chart for monthly signups: Jan 120, Feb 155, Mar 171, Apr 205. Add a short trend summary.
```
## Create a QR code
Use QR codes for public links and non-sensitive text.
Do not put passwords, verification codes, private keys, payment details, or confidential information in a QR code. Anyone who can scan the image can read what it contains.
## Ask for pictures from web search
When a search-backed answer is about something you can see, such as a place, landmark, building, product, artwork, animal, vehicle, or dish, Babbily can add a strip of up to three pictures above the answer.
* **First question in a chat:** Babbily decides whether pictures help. Asking what something looks like also lets Babbily decide, even later in a chat. A chat continued from a News article or Finance page starts fresh for this purpose.
* **Follow-up questions:** Babbily adds search pictures only when you ask to see them.
* **Direct requests:** Say "show me photos of," "pictures of," or "diagram of" to ask for pictures at any point.
```text theme={null}
Show me photos of the Sagrada Familia interior.
```
Pictures appear once the written answer starts, not before it. Babbily filters out logos, icons, ads, tiny thumbnails, and pictures whose captions do not match your question, so some searches show no pictures even when you ask. Open a cited source when you want to see more from that page.
Search pictures are different from generated images. A request such as "create an image of a cat" uses [image generation](/features/image-and-video-generation) instead. Deep Research reports cite text sources and do not include search pictures.
## Review sources, media, and Canvas artifacts
Search and research replies can include citations so you can open source material directly. Search pictures appear in a strip above the answer once the written reply starts, and a source with a picture may show a small preview in the sources list. Deep Research streams its planned report into Canvas, tells you when source coverage is partial, and preserves an incomplete report if writing ends early. Generated images and videos may include save, download, preview, or fullscreen controls depending on the output and your account state. Canvas artifacts may include editing, version, export, download, or plugin actions depending on the artifact type.
Learn when to use search and deep research for source-backed answers.
Create media with better prompts and understand usage impact.
Edit generated work in Canvas and export or send it when supported.
## Troubleshooting visual replies
If a visual reply does not appear:
* Ask for the format explicitly.
* Ask to see pictures directly, such as `Show me photos of this landmark`, if a search reply has none.
* Switch to a model that supports the visual or tool you need.
* Use **Manual** mode and choose the relevant tool.
* Provide cleaner labels, values, locations, or source files.
* Ask for a Canvas artifact if you want an editable draft, table, code file, or structured output.
* Check your API usage budget before running image or video generation again.
* Verify important outputs against the original data or cited source.
Babbily may answer in text when the requested visual or Canvas artifact is unsupported, the data is not structured enough, or the selected model cannot use the needed tool.
# Voice dictation
Source: https://docs.babbily.com/features/voice-dictation
Use your microphone to dictate chat prompts or translation text.
Voice dictation lets you speak instead of typing when your browser or device supports it.
## Where you can use dictation
You can use voice dictation in:
* Chat.
* Translate.
Look for the microphone button near the message box or source text field.
## Dictate text
Click the microphone button.
If your browser or device asks for permission, allow microphone access.
Babbily listens and adds the transcript to the text field.
Check the text before you send a chat message or use a translation.
## Tips for better dictation
* Use a quiet environment when possible.
* Speak in complete phrases.
* Review names, numbers, and technical terms before sending.
* In Translate, choose the source language when auto-detect is not enough.
## Troubleshooting
If dictation does not start:
* Check microphone permission for your browser or app.
* Make sure another app is not using your microphone.
* Try refreshing Babbily.
* Type or paste text if your current browser does not support voice input.
Review dictated text before sending it. Microphone transcription can misunderstand names, numbers, accents, or specialized terms.
# Admin rollout checklist
Source: https://docs.babbily.com/guides/admin-rollout-checklist
Plan a team rollout for Babbily with access, billing, privacy, plugins, usage, training, and support readiness.
Use this checklist when you are introducing Babbily to a team or expanding access beyond a small pilot.
Set up a team workspace, invite members, and understand team usage.
Review team, SSO, support, privacy, and procurement questions.
## Before inviting teammates
Choose the person who should manage billing, membership, and team setup.
Review included API usage budget, team seat requirements, and billing expectations.
Start with the roles or departments that have clear Babbily use cases.
Tell teammates what types of data they may use in Personal and team workspaces, and what they should keep out.
Choose whether useful team-chat context should be reusable by all members. Document what information is appropriate before the owner turns the setting on.
Decide which connected apps are appropriate for the rollout and who should authorize them.
## Recommended launch sequence
Create the team, confirm the team name, and review owner-only settings in **Settings > Team**.
Invite teammates by email. Remind them to use the invited email address and complete onboarding before accepting the invitation.
Send teammates the [quickstart](/quickstart), [Teams](/features/teams), [prompting patterns](/guides/prompting-patterns), and [role-based getting started](/guides/role-based-getting-started) pages. Ask them to check the workspace pill before starting work.
Explain that AI work uses the plan's API usage budget and that heavier work such as research, media generation, and long model runs may use more.
Share guidance for memory, files, plugins, sharing, and sensitive information before teammates start uploading or connecting apps.
Tell teammates who owns internal questions and when they should contact Babbily support.
## Suggested internal announcement
```text theme={null}
We are rolling out Babbily for AI chat, research, file work, and approved connected-app workflows.
Start here: [Quickstart]
Use Babbily for: [your approved use cases]
Do not use Babbily for: passwords, private keys, payment card numbers, government identifiers, or confidential data you are not allowed to process.
Questions: [internal owner or support path]
```
## Ongoing admin habits
* Review team members and pending invitations regularly.
* Cancel invitations that are no longer needed.
* Remind users to remove connected-app access when roles change.
* Review shared links and Library items when work changes ownership.
* Review team-wide shares and the Shared team memory setting.
* Revisit usage if the team approaches its plan budget.
* Update internal guidance when approved use cases change.
Review members, invitations, shared links, plugin access, billing, and usage when team access changes.
# Glossary
Source: https://docs.babbily.com/guides/glossary
Learn the Babbily terms you will see across chat, memory, tools, plugins, teams, and billing.
Use this glossary when a Babbily setting, guide, or support reply uses a term you want to understand.
## Everyday terms
A conversation with Babbily. Chats can include messages, files, tools, generated media, citations, and follow-up questions.
The AI system selected for a reply. Models can differ in speed, reasoning, file support, media support, context size, and API usage budget impact.
The amount of recent conversation, file content, and tool output the selected model can consider for one reply. Long chats can fill the context window, and Babbily may compact older context so the thread can continue.
The place to find saved work such as threads, images, audio, videos, and translations. Available item actions can include reopen, rename, share, download, or delete.
A Personal or team context that keeps chats, Library items, memory, and plan context organized. The workspace pill shows where you are working.
A personal or team workspace container that keeps related chats, instructions, knowledge files, and project memory together.
A paid-plan feature that runs a saved prompt manually or on a schedule and leaves each result in a new chat.
A chat mode for conversations you do not want saved in normal history or Library. Temporary chats do not use or update Babbily memory, and exiting deletes the temporary conversation.
## Personalization
A readable view of saved context Babbily may use to personalize future replies. You can review it in **Settings > Memory**.
A source or note that helps explain saved memory. You can review memory citations from **Settings > Memory** and remove eligible items.
Older saved memories awaiting your approval. They are excluded from saved-memory use in replies until approved. You can approve or forget them in **Settings > Memory**.
A limited search of eligible earlier conversations in **Auto** mode with memory enabled. It is separate from saved memory and does not cover every conversation or workspace.
An owner-controlled team setting that lets useful facts from team-workspace chats become reusable context for current team members. It does not share anyone's personal Memory Profile.
Reusable instructions that help Babbily follow a workflow. You can turn skills on or off, search verified skills, add markdown skills, and trigger skills with slash commands.
## Tools and apps
A capability Babbily can use beyond a model-only reply, such as web search, deep research, QR codes, weather, image generation, video generation, or connected-app work.
A multi-step research tool that plans a broad question, gathers and checks sources, and writes a cited report into Canvas. A run can take several minutes. If evidence gathering reaches its limit, Babbily tells you coverage is partial; if writing reaches its limit, Canvas keeps the written portion and labels it incomplete.
A source-aware tool for live quotes, historical company questions, earnings, filings, peers, and other financial research. Chats started from Finance use their page context and follow separate Auto, Manual, and Off rules.
Tool modes in the composer. **Auto** lets Babbily choose available tools, **Manual** lets you pick one, and **Off** keeps the reply model-only.
A connected app Babbily can use after you authorize access, such as Gmail, Google Drive, Slack, GitHub, Notion, Outlook, Stripe, or other supported apps.
A plugin action that asks for your confirmation before it runs. The approval card appears inline in the conversation. Review sends, updates, workflow changes, and billing or customer-data actions before choosing **Allow**.
## Account and billing
The monthly plan budget for supported AI work Babbily runs for you. Usage can include chat replies, model work, media generation, translation, and some tool or file work. Some features may use separate limits or availability rules.
The person who manages a team workspace, invitations, member removal, plan changes, billing, and team usage controls where available.
A person who has accepted a team invitation and can use the team plan. Team members may not see owner-only billing or membership controls.
A link to a shared thread. Public shared threads can be viewed by people with the link and may be discoverable when public search visibility is enabled.
Read-only access to a supported team-workspace chat or saved media item for specific teammates or everyone in the team. Access depends on current team membership.
Start with the main Babbily workflow.
Understand why behavior can vary by model, tool, file, plugin, plan, and usage state.
# Prompting patterns
Source: https://docs.babbily.com/guides/prompting-patterns
Write Babbily prompts that produce clearer, more useful answers.
Good prompts tell Babbily what you are trying to accomplish, what context matters, and how you want the answer shaped.
For work you repeat, reuse a consistent prompt structure and save stable context such as your role, audience, tone, and formatting preferences to memory. Keep task-specific instructions in the prompt so you can adjust them for each chat.
## The simple pattern
Use this structure when you want a reliable answer:
```text theme={null}
Goal: [what you want]
Context: [what Babbily should know]
Audience: [who this is for]
Constraints: [tone, length, sources, tools, files, or style]
Format: [bullets, table, memo, plan, checklist, draft, chart]
```
Example:
```text theme={null}
Goal: Draft a customer update.
Context: We fixed the issue and monitoring is stable.
Audience: Non-technical customer admins.
Constraints: Calm, concise, no blame, include next steps.
Format: Email draft with subject line.
```
## Ask for the output shape
If you do not specify a format, Babbily chooses one. Ask for the format you need:
* `Return this as a table.`
* `Give me a 5-step checklist.`
* `Write this as a customer email.`
* `Summarize this as risks, decisions, and next steps.`
* `Show this as a bar chart if the data is clear enough.`
## Tell Babbily what to use
Name the context source when it matters.
Examples:
* `Use the attached file.`
* `Use web search and cite sources.`
* `Show me photos of the venue.`
* `Use Gmail to find the latest customer thread.`
* `Use my Memory Profile only if it helps.`
* `Do not use tools for this reply.`
## Improve an answer with follow-ups
Useful follow-ups:
* `Make it shorter.`
* `Add examples.`
* `Explain the trade-offs.`
* `Turn this into a checklist.`
* `Make it more customer-friendly.`
* `What assumptions did you make?`
* `What should I verify before using this?`
## Prompt recipes
```text theme={null}
Draft a [format] for [audience]. The goal is [goal]. Use a [tone] tone. Include [required details]. Avoid [things to avoid].
```
```text theme={null}
Research [topic]. Use current public sources, cite them, and return a table with findings, source links, confidence, and open questions.
```
```text theme={null}
Analyze the attached file for [goal]. Return a summary, risks, important details, and recommended next actions.
```
```text theme={null}
Use [plugin] to find [context]. Summarize [desired result]. Ask before sending, updating, or changing anything.
```
```text theme={null}
Create [asset type] for [audience/use]. Style: [style]. Include [elements]. Avoid [constraints]. Size/aspect ratio: [size].
```
## When answers miss the mark
Try:
* Restating the goal.
* Adding the audience.
* Asking for a different format.
* Providing an example of the style you want.
* Saying what to ignore.
* Starting a new thread if old context is steering the answer.
* Switching models when the task needs a different capability.
Start from complete workflows for files, research, plugins, media, and translation.
Choose the right model for speed, reasoning, files, tools, media, and context.
# Role-based getting started
Source: https://docs.babbily.com/guides/role-based-getting-started
Pick the Babbily setup path that matches how you work.
Different people get value from Babbily in different ways. Use the path that sounds closest to your work, then adjust it as your workflow grows.
## If you write, summarize, or plan
Ask Babbily to draft, rewrite, summarize, compare, or plan. Include your audience and desired format.
Add memory for tone, format, role, units, or recurring audience details you want reused across chats.
Use a consistent prompt structure for repeatable work such as weekly updates, customer emails, or review checklists. Your saved memory can supply stable tone, audience, and formatting preferences, so the prompt can focus on the current task.
Try:
```text theme={null}
Draft a concise customer update for non-technical admins. Include what changed, what they need to do, and who to contact.
```
## If you research decisions
Use **Manual > Web search** when you need current public sources.
Use **Manual > Deep research** when you need a multi-source brief, comparison, or landscape. Follow the plan and report in Canvas, then review its citations and completion status.
Open sources before relying on pricing, policy, legal, medical, financial, or fast-changing information.
Try:
```text theme={null}
Do deep research on this market and return a sourced brief with key players, risks, and open questions.
```
## If you work with files
File controls appear only when the selected model supports the file type you want to use.
Add one or a few relevant files first. Long files can fill the context window quickly.
Ask for a summary, risks, table, rewrite, action list, comparison, or decision memo.
Try:
```text theme={null}
Analyze this document for executive readers. Return a summary, open decisions, risks, and next steps.
```
## If you manage ongoing or recurring work
Keep related chats, instructions, knowledge files, and project memory together in the workspace where the work belongs.
On a paid plan, use **Scheduled** for briefings, reviews, or monitoring that should run without you present.
Scheduled Tasks can read from selected Plugins, but unattended runs cannot stop for approvals or perform write actions.
## If you use connected apps
Open **Settings > Plugins** and authorize the app you want Babbily to use.
Name the app, person, folder, project, channel, repository, or time window when it matters.
Approve plugin actions only after checking recipients, destinations, content, and side effects.
Try:
```text theme={null}
Use Gmail to find the latest thread with this customer and summarize open questions, blockers, and next steps.
```
## If you manage a team
Team plans use seat-based billing with a three-person minimum and include team workspace, billing, member, and usage controls where available.
Explain when members should use **Personal** or the team workspace, then decide whether the owner should enable **Shared team memory**.
Use **Settings > Team** to invite people and review pending invitations.
Check accepted members, pending invitations, team shares, Shared team memory, public shared chats, plugin access, billing, and API usage budget.
Organize ongoing work in a personal or team workspace.
Automate recurring prompts and review their run history.
Follow step-by-step workflows for files, research, connected apps, media, and translation.
Structure reusable instructions for common tasks.
Fix authorization, status, permission, and approval issues.
Review team, billing, privacy, and procurement questions.
# Task recipes
Source: https://docs.babbily.com/guides/task-recipes
Use these customer-friendly recipes to get useful work done faster in Babbily.
These recipes show common ways to use Babbily. Treat them as starting points, then adjust the prompt, model, files, tools, or plugins for your work.
Attach a file and ask for a summary, table, risks, or next steps.
Use search or deep research for current, source-backed answers.
Turn a chat into a memo, brief, document, or slide outline.
Ask Babbily to summarize email, documents, project work, or app context.
Keep chats, instructions, knowledge files, and memory in a Project.
Create a paid Scheduled Task and review each run in a new chat.
## Analyze a file
Use this when you want Babbily to summarize, compare, extract, rewrite, or turn an attachment into action items.
Start a new chat or open an existing thread where the file belongs.
Pick a model that supports the file type you want to use.
Add one or a few relevant files. Long files can fill the context window, so include the most useful material first.
Name the result you want: summary, risks, table, action items, comparison, rewrite, or plan.
Ask follow-up questions while the file context is still useful.
```text theme={null}
Summarize this document for an executive audience. Include open risks, decisions needed, and a five-item action list.
```
Do not upload secrets, private keys, payment details, or files you are not allowed to use in Babbily.
## Research with citations
Use this when a reply needs current information, public sources, or a broader research report.
Use **Auto** for most research. Use **Manual > Web search** for focused current lookup, or **Manual > Deep research** for broader multi-source work.
Tell Babbily what to research and what sources or time window matter. Add "show me photos of" when you want pictures of the subject with the answer.
Ask for a brief, comparison, timeline, pros and cons, source table, or recommendation. Deep Research writes the full report into Canvas.
Open cited sources before relying on important conclusions.
If chat says the coverage is partial or Canvas labels the report incomplete, review what it covers and narrow the next request to the missing sections.
```text theme={null}
Search current public sources and compare these three vendors. Return a table with pricing signals, strengths, risks, and citations.
```
If a selected plugin action needs approval during Deep Research, review the inline approval card. The Canvas report keeps the work gathered before the pause.
## Turn a chat into a document or slide outline
Use this after a useful chat, brainstorm, research run, or meeting-prep thread.
Reopen the thread you want to turn into a deliverable.
Name the format: memo, brief, meeting notes, project plan, document outline, or slide outline.
If you use Google Docs, Google Slides, Notion, or another supported app, connect it first and ask Babbily to prepare the content there when available.
Review any draft, send, update, or workspace action before approving it.
```text theme={null}
Turn this thread into a six-slide outline for a customer strategy review. Include slide titles, bullets, and speaker notes.
```
## Summarize emails with a Plugin
Use this when you want Babbily to work with email or calendar context you authorize.
Go to **Settings > Plugins** and connect Gmail or Outlook.
Tell Babbily what to find, who or what to search for, and what output you want.
Ask for open questions, decisions, blockers, next steps, or a draft reply.
Review any send, update, or workspace action before approving it.
```text theme={null}
Use Gmail to find the latest thread with this customer and summarize open questions, blockers, and suggested next steps.
```
## Create an image or video from source media
Use this when you want a new visual asset, a variation, or a short video.
For an image variation, select an image model that supports source-image editing. The chat **Image generation** tool creates new images from text and does not edit attached images. For video, choose a supported video model or tool.
Attach required source media when the selected model asks for it, or attach optional source media when you want Babbily to use it as a reference.
Include subject, style, composition, motion, pacing, size, aspect ratio, and constraints.
Save, download, or ask for another version.
```text theme={null}
Use this image as reference and create a cleaner 16:9 version for a launch announcement. Keep the layout simple and bright.
```
## Translate and continue in chat
Use this when you want a translation and then want Babbily to help with editing, replying, or planning.
Open **Translate** from the sidebar.
Select source and target languages, or use auto-detect when available.
Paste text or use voice dictation when your browser and device support it.
Pick a tone or style chip when you want the translation to sound different.
Use **Continue in chat** to open a chat with the original and translated text as context.
```text theme={null}
Rewrite this translated reply to sound polite, concise, and customer-ready.
```
## Related guides
Keep reusable instructions and files with a set of related chats.
Run a complete prompt on a schedule or with **Run now**.
Attach documents, images, audio, and video when a chat needs file context.
Improve prompts by adding goals, context, audience, constraints, and output format.
See what each connected app can help you find, reference, draft, or update.
Ask for maps, charts, stat cards, QR codes, citations, images, and videos.
Understand how plan usage, resets, and heavier tasks work.
# Welcome to Babbily
Source: https://docs.babbily.com/index
Use Babbily to chat with leading AI models, remember important context, connect your apps, create media, and manage your workspace.
Babbily helps you work with AI from one place. You can chat with leading models, search the web, research topics, generate images and videos, connect your apps, and let Babbily remember context you choose to keep.
Create your account, send your first message, and learn where the main controls live.
Create your account, verify email, reset passwords, accept invites, and manage access.
Choose models, send messages, use attachments, and keep long conversations moving.
Create, edit, save, export, and send AI-generated work without leaving chat.
Pick the right model for speed, reasoning, files, tools, media, context, and usage.
Follow step-by-step recipes for files, research, connected apps, media, and translation.
Pick a setup path for writing, research, files, connected apps, or team administration.
Launch Babbily with team access, privacy guidance, plugins, usage, and support paths.
Learn the Babbily terms used across chat, memory, tools, plugins, teams, and billing.
Write clearer prompts with goals, context, audience, constraints, and output format.
Review what Babbily remembers and control what it can use in future chats.
See what to save to memory, what to keep in chat, and what to avoid.
Decide when Babbily can search, research, generate media, or use reusable skills.
Keep related chats, instructions, knowledge files, and project memory together.
Run recurring prompts automatically from a personal or team workspace.
Connect workspace apps and approve actions before anything is sent or changed.
Use practical examples for email, files, docs, spreadsheets, chat, project, developer, and business apps.
Ask for maps, charts, weather cards, QR codes, citations, images, and videos.
Browse current stories, review citations, set local context, and continue from an article in chat.
Review markets, company pages, watchlists, earnings, prediction markets, and source-aware finance chat.
Translate text, choose a style, use voice input, and continue in chat.
Understand your plan, monthly usage budget, reset timing, and team usage.
Understand plan changes, billing visibility, team billing, and refresh timing.
Review team, SSO, support, privacy, and procurement questions.
Prepare for vendor review with privacy, SSO, billing, support, and data-control topics.
Review privacy controls for memory, plugins, sharing, Library items, and account help.
See what you can copy, download, delete, disconnect, or request through support.
Review browser, device, permission, and accessibility support expectations.
Fix verification code, password reset, onboarding, team invite, and share-link access issues.
Troubleshoot permissions, devices, voice input, and unsupported files.
Understand what to try when a model, tool, plugin, media job, billing, or login is unavailable.
Fix plugin setup, authorization, permission, and approval issues.
Understand what can vary by model, tool, file, plugin, plan, usage, and context.
New to Babbily? Start with the [quickstart](/quickstart), then open [Memory](/features/memory) and [Tools and skills](/features/tools-and-skills) when you are ready to personalize your workspace.
# Quickstart
Source: https://docs.babbily.com/quickstart
Create your account, send your first Babbily message, and learn the main controls.
Use this guide to get from a new account to your first useful chat.
## What you need
* A Babbily account.
* A supported web browser.
* A question, task, file, or idea you want help with.
## Start your first chat
Go to Babbily and sign in. If you are new, create an account and finish onboarding.
Type a clear request in the composer. For example:
```text theme={null}
Help me draft a concise project update for this week's team meeting.
```
Use the model picker when you want a different balance of speed, reasoning, creativity, or media support. Some tools and file types appear only for models that support them.
Attach images, documents, audio, or video when you want Babbily to use file context. File support depends on the active model.
Open the composer menu and choose **Auto**, **Manual**, or **Off**. Auto lets Babbily decide when to use tools. Manual lets you choose. Off keeps the reply model-only.
## Save context with memory
If there is something you want Babbily to remember across chats, say it directly:
```text theme={null}
Remember that I prefer short answers with clear next steps.
```
Check the memory status to confirm the save. You can review and remove saved context later in **Settings > Memory**, including older items marked **Needs review**. See [memory controls](/features/memory).
Use memory for stable preferences and ongoing context. For one-time details, keep them in the current chat instead.
## Check your usage
Your plan includes a monthly API usage budget. The usage card shows how much of that budget you have used and when it resets. Open **Settings > Billing** for plan and renewal details.
## Where to go next
Learn the terms you will see across Babbily.
Pick a setup path that matches how you work.
Learn how to choose models and structure better chats.
Choose a model for speed, reasoning, files, tools, media, context, and usage.
Try common workflows for files, research, connected apps, media, and translation.
Use goals, context, audience, constraints, and formats to get better answers.
Find profile, billing, team, plugin, skill, and memory settings.
Control what Babbily remembers about you.
Learn what belongs in memory and what should stay in the current chat.
Use search, research, skills, and media tools.
Keep ongoing chats, instructions, knowledge, and memory together.
Run saved prompts automatically on a paid plan.
Connect apps and choose the context Babbily can use.
Troubleshoot upload, microphone, and device behavior.
Ask for maps, charts, weather cards, QR codes, citations, images, and videos.
Browse current stories, set local context, review citations, and ask follow-up questions.
Review markets, company pages, watchlists, earnings, and prediction markets.
Understand plan usage and reset timing.
See what can vary by model, plan, tool, file, plugin, usage, and context.
Translate text, choose a style, and continue in chat.
# Browser, device, and file support
Source: https://docs.babbily.com/troubleshooting/browser-device-and-file-support
Troubleshoot browser permissions, device behavior, notifications, voice input, attachments, and unsupported files.
Babbily works across modern web browsers and supported app surfaces, but some features depend on browser permissions, device capabilities, the selected model, and your account state.
## Quick checks
| Issue | What to check |
| -------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------- |
| Microphone does not start | Browser or app microphone permission, another app using the microphone, device support. |
| Plugin window does not open | Pop-up permissions for Babbily and whether the authorization window was blocked. |
| File upload is hidden | Selected model file support and your current account state. |
| File says unsupported | File type, selected model, and whether the model supports that file category. |
| Media generation asks for a file | Whether the selected media model requires source media. |
| Notifications do not arrive | **Settings > Account > Notifications**, browser or device notification permission, supported browser, and whether you were already active in Babbily. |
| Passkey is unavailable | Supported web browser, device passkey support, password manager or security key availability, and account state. |
| Page looks stale | Refresh Babbily after changing permissions, plans, models, or connected apps. |
## File support basics
File support is model-dependent. Depending on the selected model, Babbily can show upload controls for:
* Images.
* Documents.
* Audio.
* Video.
Document uploads from the file picker can include common text and document formats such as `.txt`, `.js`, `.ts`, `.tsx`, `.py`, `.html`, `.css`, `.json`, `.csv`, `.md`, `.pdf`, `.doc`, and `.docx`.
Image uploads generally work best as PNG, JPG/JPEG, or WebP when the model needs to read the image. Other image, audio, or video files may appear in the picker when the selected model supports that broad category, but not every model can read every format.
The model picker and composer controls are the best source of truth for what your current chat can attach.
## If a file will not upload or send
Try these steps:
Choose a model that supports the file category you need.
Add the most relevant file before adding more context.
If a document or image fails, try a common format such as PDF, DOCX, TXT, CSV, PNG, JPG, or WebP.
Large files can fill the context window or fail if the selected model cannot handle them.
Use a clean thread when older files or instructions are no longer relevant.
## Voice dictation
Voice dictation depends on browser support, device support, microphone permission, and account state. Some app surfaces may use a transcription flow when browser speech input is unavailable.
If dictation fails:
* Allow microphone access.
* Close other apps that may be using the microphone.
* Try a shorter recording.
* Refresh Babbily after changing permissions.
* Type or paste text if your current browser or device does not support dictation.
Review dictated text before sending it. Transcription can misunderstand names, numbers, accents, or technical terms.
## Notifications
Completion notifications depend on browser support, device support, notification permission, and your Babbily setting.
If notifications do not appear:
* Open **Settings > Account > Notifications** and confirm **Chat completion notifications** is on.
* Allow notifications for Babbily in your browser or device settings.
* On iPhone or iPad, use **Share > Add to Home Screen**, then open Babbily from the Home Screen and enable notifications there.
* Use Babbily in a supported browser. Push notifications may be unavailable in some embedded, private, or restricted browser surfaces.
* Refresh Babbily after changing notification permissions.
* Check whether you were already looking at Babbily when the run finished. Babbily may skip a notification when you are already active in the app.
## Passkeys
Passkeys are currently available on supported web browsers. Use another sign-in method in the native app if passkeys are unavailable.
If passkeys do not appear or a prompt fails:
* Refresh Babbily in a supported browser.
* Confirm your device, password manager, or security key supports passkeys.
* Sign in another way and add, rename, or remove passkeys from **Settings > Account**.
## Plugin authorization windows
Some plugin setup flows open a separate authorization window and then return you to Babbily.
If the window does not open:
* Allow pop-ups for Babbily.
* Choose **Authorize** again.
* Complete the authorization flow in the app you are connecting.
* Return to **Settings > Plugins** and refresh.
Learn how to attach files and get better file results.
Use the microphone in Chat and Translate.
Fix plugin setup, authorization, status, and permissions.
Understand why behavior can vary by model, file, plan, usage, and account state.
# Common issues
Source: https://docs.babbily.com/troubleshooting/common-issues
Fix common problems with chat, tools, plugins, memory, usage, and billing.
Use this page when something in Babbily does not behave the way you expected.
Understand what can vary by model, tool, file, plugin, plan, usage, and context.
Fix verification, password reset, onboarding, team invite, and share-link access issues.
Understand what to try when a feature is slow, unavailable, or temporarily limited.
## A model is missing or marked Sunsetting
Model availability changes over time. **Sunsetting** models can remain selectable during a transition; models removed from the picker can no longer be selected there.
* Refresh the page to load the current model list.
* Check [model transitions and alternatives](/features/model-capabilities#sunsetting-and-removed-models), then choose an available model for new work.
* If you are using **Translate**, check its own picker. Its model list differs from chat.
* If a **Scheduled Task** uses an unavailable model, edit the task to choose a current model and review **Recent runs**.
An older model name in saved history does not mean it is still available for new requests. A missing **Fast mode** control can also be expected: not every model supports it, including models with “Flash” in their names.
## A tool is missing
Try these steps:
Some tools appear only for models that support them. Switch models if the tool is hidden.
Use **Auto** or **Manual**. In **Off** mode, Babbily does not call tools.
Some capabilities can depend on your plan or remaining usage.
## A plugin is not ready
If a plugin shows setup, authorization, or preparation status:
* Finish the plugin setup flow.
* Refresh **Settings > Plugins**.
* Reconnect if authorization changed.
* If you do not see a disconnect option, revoke access from the connected app's own settings and contact support if you still need help.
* Wait a moment if the plugin was just connected.
See status meanings, pop-up fixes, permission checks, approval reminders, and access removal steps.
## Babbily did not search
If you expected live information, switch to **Manual** and choose **Web search** or **Deep research**. You can also ask directly:
```text theme={null}
Search the web and cite sources for the latest information about this topic.
```
## Search replies show no pictures or unhelpful ones
Web search adds pictures only when they fit. For the first question in a chat, Babbily can include pictures of a subject you can see, such as a place, product, or landmark. In follow-up questions, it adds search pictures only when you ask to see them:
```text theme={null}
Show me pictures of the Eiffel Tower at night.
```
Babbily leaves out logos, icons, ads, tiny thumbnails, and pictures whose captions do not match the search, and it shows pictures only after the written answer starts. Some searches have no usable pictures even when you ask. Deep Research reports do not include search pictures, and a request to create an image uses [image generation](/features/image-and-video-generation) instead.
## A Deep Research report stopped or is waiting
Open the report in Canvas and check its status.
* If chat says the coverage is partial or Canvas says the report is incomplete, review the usable sections and narrow a follow-up to the missing coverage.
* If an approval card appears inline in the conversation, review the plugin action and choose **Allow** or **Cancel**.
* If the report still says it is in progress after the chat shows a final result, refresh the chat once.
* If no report was created, try a narrower request, remove an unavailable plugin, switch to a supported model, or retry later.
## Memory feels wrong
Open **Settings > Memory** and review the Memory Profile and citations. Remove anything outdated. You can also say:
```text theme={null}
Forget that I prefer long answers.
```
If a memory cannot be removed from the UI, contact support.
If a save, approval, or forget says memory is temporarily unavailable, wait a moment and try again. Contact support if it keeps happening.
If the issue happened in a team workspace, check the workspace pill and open **Settings > Team** to see whether **Shared team memory** is enabled. Your personal Memory Profile and team memory are separate.
If older saved context is missing, check **Needs review** in **Settings > Memory**. Approve only the items you want to reuse. Forgetting an item does not remove its original chat history.
For earlier-conversation questions, use **Auto** mode with memory enabled and review [past-chat lookup availability](/features/memory#refer-to-past-chats).
## An earlier message will not edit
Wait for active work to finish. If the original attachment is incompatible, select a compatible model; if it is missing, send a new message with the file attached. If Babbily says the edit may have been saved, refresh before retrying.
Sending an edit removes later messages. See [editing an earlier message](/features/chat-and-models#edit-an-earlier-message).
## A thread will not move into a project
Use your own eligible saved thread and a writable project in the same workspace. Wait for the thread's current action to finish. If the thread changed elsewhere, refresh, review its current project, and try again. Temporary Chats and Scheduled Task results cannot be filed this way.
See [adding an existing thread](/features/projects#add-an-existing-thread).
## Work is missing after switching workspaces
Chats, Library items, and memory stay with the workspace where they were created.
* Open the workspace pill or account menu.
* Switch to **Personal** or the team where you created the item.
* Reopen Library or the relevant feature.
* Ask the team owner whether your membership or the team plan changed.
If a team workspace says **Plan inactive — read-only**, the owner must restore its plan before members can create new work there.
## A Project is missing or read-only
* Switch to the workspace where the Project was created.
* Check whether it is archived.
* Confirm team membership and team-plan status.
* Turn off Temporary Chat before starting a Project chat.
* Refresh if a knowledge file is still processing.
Review project access, knowledge files, memory, and lifecycle behavior.
## A Scheduled Task did not run
Open **Recent runs** and review the latest status. Confirm the workspace still has a paid plan and available API usage budget. If the task says **Plugin reselection required**, reconnect the app in **Settings > Plugins**, edit the task, choose its current Plugin access, and save it before retrying.
Review plan checks, workspace limits, Plugin access, and run failures.
## Usage looks delayed
The usage card may take a short time to refresh after model, tool, image, or video work. Wait a moment and refresh the page. If usage still looks wrong, check **Settings > Billing**.
## A chat is near the context limit
Start a new thread if the topic changed or the context indicator is full. Babbily may compact older context, but very long threads can still become harder to steer.
## Image or video generation failed
Try a shorter, clearer prompt. If you attached source media, confirm that the selected model supports it. Also check your usage budget before trying again.
## Translate is not updating
Try these steps:
* Make sure a translation model is available in the model picker.
* Check that the source text is not empty.
* Wait a moment after you stop typing.
* Try a shorter section of text.
* Check your API usage budget if translation stops with a usage message.
## Voice dictation is unavailable
Voice dictation depends on your browser, device, microphone permission, and account state.
If the microphone button does not work, allow microphone access, close other apps that may be using the microphone, refresh Babbily, or type the text manually.
Troubleshoot microphone, pop-up, upload, and device support issues.
## News is empty or not local
If News does not show what you expect:
* Try another category.
* Update the location field on the News page.
* Use the current-location control if your browser or device allows it.
* Try a broader location, such as a country instead of a city.
* Open the original sources before relying on a story for important decisions.
## Finance data is unavailable
If Finance does not show the market, company, or historical data you expected:
* Refresh the Finance page.
* Search by company symbol when a company name is ambiguous.
* Try another Finance tab, such as **US Markets**, **Crypto**, **Earnings**, or **Predictions**.
* In a Finance thread, use **Auto** to run supported Finance research automatically, keep **Finance research** enabled in **Manual**, or use **Off** for a model-only reply.
* Use a precise company symbol or full company name when the issuer is ambiguous.
* Open cited stories or source links directly before relying on important details.
* Wait and retry if the page says Finance data is temporarily unavailable.
* Use **Retry** after a temporary research failure; the thread keeps its Finance context and tool setting.
If Babbily says a requested metric is not reported separately, ask for one of the related reported figures it suggests instead of retrying the same unavailable metric.
## Billing controls are missing
Billing controls can depend on your role. Team members may not see owner-only billing actions. Ask the team owner to manage subscription changes.
## Login or invite access fails
Use the latest verification code or invitation email, confirm the account email, and finish onboarding if prompted. If the team still does not appear after accepting an invite, ask the team owner to check the invited email and pending invitations.
Fix sign-in, verification code, password reset, onboarding, and team invitation issues.
If the issue persists, contact support with the page you were on, what you expected, and what happened.
See what to include, what not to send, and which details help support investigate faster.
# Contact support
Source: https://docs.babbily.com/troubleshooting/contact-support
Learn what to include when you need help with Babbily and what sensitive details to leave out.
Contact support when you cannot resolve an issue from the product UI or the troubleshooting guides.
Use the support site for account, billing, plugin, usage, memory, and product issues.
## What to include
The fastest support requests include enough context to reproduce or investigate the issue.
| Include | Why it helps |
| ------------------------------------- | --------------------------------------------------------- |
| Account email | Helps support find the correct account. |
| Page or feature | Shows where the issue happened. |
| What you expected | Makes the intended outcome clear. |
| What happened instead | Shows the visible failure or confusing result. |
| Approximate date, time, and timezone | Helps narrow the event window. |
| Browser, device, and operating system | Helps identify environment-specific issues. |
| Plan or team role | Helps explain billing, usage, and permission differences. |
| Error text or screenshots | Shows the exact customer-visible state. |
## What not to send
Do not send:
* Passwords.
* Verification codes.
* Private keys.
* API keys.
* Payment card numbers.
* Government identifiers.
* OAuth tokens or raw connected-app credentials.
* Confidential files or chat content unless support specifically requests them through an approved support channel.
Hide sensitive details in screenshots before sending them to support.
## Common support topics
Include the account email, plan name, team role, billing page you were viewing, what you expected, and any customer-visible error text. Team members should also ask the team owner to check owner-only billing controls.
Include the plugin name, whether it shows connected or pending, the action you tried, whether approval was requested, and whether your role in the connected app may limit access.
Include whether the issue is about your Memory Profile, a memory citation, a chat reply that used outdated context, or a memory you cannot remove from **Settings > Memory**.
Include the plan, approximate time of the model/tool/media action, what the usage card showed before and after, and whether you are part of a team workspace.
Include the model, tool mode, prompt, file type if you attached source media, and whether the result failed, timed out, or returned text instead of the visual format you expected.
## Useful links
Try quick fixes for tools, plugins, memory, usage, media, translation, voice, news, finance, and billing.
Review privacy, memory, plugin, sharing, Library, and account controls.
Learn what to try when a feature is temporarily slow, unavailable, or limited.
Understand what you can remove yourself and what may require support.
# Known limits and expectations
Source: https://docs.babbily.com/troubleshooting/known-limits-and-expectations
Understand why Babbily behavior can vary by model, tool, file, plugin, plan, usage, and context.
Babbily combines models, tools, memory, plugins, files, and media features. Some behavior can vary based on the selected model, your plan, your usage state, and the access you authorize.
## Capabilities can vary
| Area | What can vary | What to try |
| ---------------- | ------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------- |
| Models | Availability, sunsetting status, speed, reasoning, context size, file support, tools, and media. | Check [current model transitions](/features/model-capabilities#sunsetting-and-removed-models) and select an available alternative. |
| Tools | Availability by model, plan, tool mode, and usage. | Use **Auto** or **Manual** and check your usage. |
| Files | Supported file types, size behavior, and how much context fits. | Attach only the most relevant files first. |
| Visual replies | Maps, charts, QR codes, media, and citations may fall back to text. Search pictures appear only when they fit the question. | Ask for the format explicitly, ask to see pictures directly, and provide cleaner data. |
| Canvas | Artifacts, export options, and plugin actions can vary by model, artifact type, and plan. | Ask for a Canvas artifact directly or try a model with stronger tool support. |
| Deep Research | Source coverage, connected-app access, run time, and report completeness can vary. | Review the chat summary and sources in Canvas; narrow a follow-up if coverage is partial or the report is incomplete. |
| News and Finance | Current stories, market data, company pages, earnings, and prediction markets can vary by source availability and market state. | Refresh, try a broader category or market tab, and verify original sources. |
| Plugins | Access depends on authorization and your role in the connected app. | Reconnect, check app permissions, or choose the plugin manually. |
| Projects | Workspace, archive state, team-plan access, knowledge limits, and Temporary Chat affect what you can add or edit. | Switch to the correct workspace, turn off Temporary Chat, and check whether the Project is read-only. |
| Scheduled Tasks | Paid-plan access, workspace limits, API usage budget, model availability, and Plugin readiness affect each run. | Open **Recent runs**, restore access, or edit the task before retrying. |
| Workspaces | Chats, Library items, and memory stay in the workspace where they were created. | Check the workspace pill and switch to **Personal** or the intended team. |
| Memory | Personal, team, and project memory have separate scopes; older items may need approval. Past-chat lookup has additional limits. | Review [memory controls and lookup availability](/features/memory). |
| Usage | Usage cards can take a short time to refresh after AI work. | Wait a moment, refresh, and check **Settings > Billing**. |
## Search and research
Search, Deep Research, News, and Finance can summarize public information and cite sources, but they can still miss sources, encounter blocked pages, or return incomplete results for fast-changing topics. When Deep Research ends early, chat identifies partial source coverage or Canvas preserves and labels the incomplete written report.
Web search adds pictures only for a subject you can see or when you ask to see them, and only pictures that pass Babbily's relevance checks. Deep Research reports do not include search pictures.
For important decisions:
* Review citations.
* Open original sources.
* Check authoritative sources directly.
* Ask Babbily to list assumptions and confidence.
## Long chats
Long threads can become harder to steer. Babbily may compact older context so a thread can continue, but old assumptions can still influence the conversation.
Start a new thread when:
* The topic changed.
* The context indicator is near the limit.
* You want a clean answer.
* Earlier files or instructions are no longer useful.
## Canvas artifacts
Canvas artifacts work best for substantial drafts, tables, code, image briefs, and structured outputs. Simple answers may stay in chat.
Export options depend on the artifact type. Plugin actions depend on the connected app, your permissions, and plugin readiness. If a structured artifact is still preparing, wait for the Canvas panel to finish loading before editing or exporting it.
## Plugins
Plugin results depend on:
* Whether the plugin is connected.
* Your role in the connected app.
* App permissions and workspace settings.
* Whether the requested item exists.
* Whether an action requires approval.
If a plugin answer seems incomplete, make the request more specific and confirm the connected app can access the item.
## Media generation
Image and video generation can take longer than text replies and usually uses more API usage budget. Results can vary by prompt clarity, source media, model support, plan, usage state, and safety checks.
Review generated media before using it publicly.
## When to contact support
Contact support if:
* A control that should be visible is missing after refreshing.
* Usage or billing still looks wrong after a short wait.
* You cannot remove data from available controls.
* A plugin stays stuck after reconnecting.
* A feature repeatedly fails with the same customer-visible error.
Review project scope, knowledge limits, memory, and read-only states.
Review plan, usage, Plugin, and run-history behavior.
Try quick fixes for tools, plugins, memory, usage, media, translation, voice, news, finance, and billing.
Troubleshoot permissions, devices, voice input, and unsupported files.
Fix plugin setup, authorization, permission, and approval issues.
Learn what to include when you need help.
# Login and invite troubleshooting
Source: https://docs.babbily.com/troubleshooting/login-and-invite-troubleshooting
Fix common sign-in, passkey, verification code, password reset, onboarding, and team invitation issues.
Use this page if you cannot access your account, complete onboarding, or accept a team invitation.
Review the normal sign-up, sign-in, passkey, password reset, and team invite flows.
## Quick checks
| Issue | Try this first |
| ----------------------------- | ---------------------------------------------------------------------------------- |
| Code did not arrive | Check spam or promotions folders, then request a new code. |
| Code is not accepted | Use the newest code and enter it exactly as shown. |
| Account not found | Confirm the email address and sign-in method you used before. |
| Account already exists | Switch from **Sign Up** to **Log In**. |
| Passkey sign-in is missing | Use a supported web browser and add a passkey from **Settings > Account** first. |
| Passkey prompt fails | Sign in another way, then refresh and manage passkeys from **Settings > Account**. |
| Invite link does not work | Ask the team owner to confirm the invited email and send a fresh invite. |
| Share link sends you to login | Sign in, then open the share link again. |
## Verification code problems
Verification and password reset codes are time-sensitive. If a code fails:
* Use the most recent email from Babbily.
* Check that you are entering the code for the same email address shown on the page.
* Request a new code if the original code is old.
* Avoid copying extra spaces before or after the code.
If codes are not arriving, check spam filters and ask your email administrator whether automated messages are being filtered.
## Password reset problems
If the reset page says a reset code is required, restart from **Forgot your password?** on the sign-in page. This helps Babbily verify which email address should receive the reset code.
If your new password is not accepted, follow the password requirements shown on the page and make sure the confirmation field matches.
If you normally sign in with a social sign-in option, use that option again. A password reset may not apply unless you added email and password sign-in to the account.
## Passkey problems
Passkeys depend on your browser, device, operating system, password manager, security key, and account state.
If passkey sign-in does not work:
* Confirm you are using Babbily in a web browser. Use another sign-in method in the native app if passkeys are unavailable.
* Make sure you already added a passkey from **Settings > Account**.
* Follow the browser, device, password manager, or security key prompt.
* Refresh the page after adding or removing passkeys.
* Sign in with another available method if your current browser or device cannot use passkeys.
If a passkey is no longer available, sign in another way and remove it from **Settings > Account**.
## Team invitation problems
If a team invitation does not work:
Older invitation emails may no longer be valid after a team owner sends a new one.
The invitation is intended for the email address the team owner entered.
New users may need to finish onboarding before Babbily can add them to the team.
If the link is expired, missing, or for the wrong email, ask the team owner to send a new invite.
## You can sign in but do not see the team
Try these checks:
* Open your account menu and check the **Workspaces** list.
* Open **Settings > Team** and refresh the page.
* Confirm you accepted the invite with the same email the owner invited.
* Ask the team owner whether the invite is still pending, accepted, cancelled, or sent to another email.
* Ask the owner to remove duplicate or outdated invitations if needed.
## When to contact support
Contact support if:
* You cannot receive any verification or reset messages.
* Your invite was accepted but team access is still missing.
* You are locked out of the account email.
* You need help with account deletion or data you cannot remove yourself.
Do not send passwords, verification codes, private keys, payment details, or raw connected-app credentials to support.
# Plugin troubleshooting
Source: https://docs.babbily.com/troubleshooting/plugin-troubleshooting
Fix Plugin setup, authorization, permission, readiness, approval, and incomplete-result issues.
Use this guide when a plugin is missing, stuck, disconnected, or returning less context than expected.
## Plugin status meanings
| Status | What it usually means | What to try |
| ------------------------- | ------------------------------------------------------------------------------- | --------------------------------------------------------- |
| Not connected | The app has not been authorized for your account. | Choose **Connect** and complete setup. |
| Waiting for authorization | The hosted authorization flow is still open or has not finished. | Finish the app flow, or choose **Continue** to reopen it. |
| Connected | Babbily can use the Plugin when the model, tool mode, and permissions allow it. | Ask for the app by name or choose it from the composer. |
| Cleanup in progress | Babbily is still disconnecting the app. | Wait a moment, then refresh **Settings > Plugins**. |
| Disconnected | The previous connection is no longer active. | Choose **Reconnect** if you want to use it again. |
| Authorization expired | The connected app authorization expired. | Choose **Reconnect** and authorize access again. |
| Authorization revoked | Access was revoked in Babbily or the connected app. | Choose **Reconnect** if you want to restore access. |
| Needs attention | Setup or connection verification did not complete. | Refresh, reconnect, or contact support if it repeats. |
## If setup does not open
Allow pop-ups for Babbily, then choose **Connect**, **Continue**, or **Reconnect** again.
Sign in to the connected app and approve the permissions you are comfortable granting.
Go back to **Settings > Plugins** and wait for the status to refresh.
Ask for the connected app by name, choose it from the composer, mention it with `@`, or use `/` when plugin-specific commands are available.
## Reconnect after the Plugin migration
Babbily retired former Smithery connections when Plugins moved to Composio. Credentials were not transferred between providers.
If an app appears under **Previous connection history**, choose the current Plugin in **Settings > Plugins** and complete the new authorization flow. An older Scheduled Task may also show **Plugin reselection required**. Edit that task, choose the connected Plugins it may use now—even if the choice is none—and save it before running or resuming the task.
## If Babbily cannot find something
Plugin answers depend on what your connected-app account can access.
Try:
* Name the app, folder, channel, project, repository, board, person, or time window.
* Confirm you can open the item directly in the connected app.
* Check whether your app role has permission to view it.
* Use **Manual** mode and choose the plugin directly.
* Reconnect if your app permissions recently changed.
## If an action needs approval
Some plugin actions ask for approval before running. Review the content and destination before approving:
* Recipients and subject lines.
* Documents, sheets, slides, pages, tasks, or records being changed.
* Billing, customer, payment, or support context.
* Automations or workflow changes.
The approval card appears inline in the conversation beside the work that requested it. Choose **Allow** to run the action or **Cancel** to decline it. During Deep Research, canceling also ends the paused run and keeps the sources gathered before the pause in Canvas.
If Deep Research is waiting for approval, the Canvas report keeps the work gathered so far. After you choose **Allow**, give the same report time to continue. If it does not resume, refresh the chat once and check whether the plugin is still connected.
Do not approve a plugin action unless you understand what it will do.
In a chat that uses shared project instructions, Babbily may ask for approval even when a Plugin only reads information. Review the app and requested access before choosing **Allow**.
## If you want to remove access
Use **Settings > Plugins** to disconnect apps where controls are available. If you cannot remove access from Babbily, revoke access in the connected app's own settings and contact support if you still need help.
Removing plugin access does not delete content inside the connected app.
Connect apps and understand approval-required actions.
See what each plugin can help you find, reference, draft, or update.
Review plugin privacy and access controls.
Learn what to include when a plugin still fails.
# Service status and degraded behavior
Source: https://docs.babbily.com/troubleshooting/service-status-and-degraded-behavior
Understand what to do when Babbily, models, tools, plugins, media generation, billing, or login are temporarily unavailable.
Sometimes a feature can be temporarily slow, unavailable, or limited even when the rest of Babbily is working. This page helps you decide what to try and what to include if you contact support.
See what to include and what sensitive details to leave out.
## Signs of degraded behavior
* Messages take much longer than usual.
* A model, tool, or plugin is temporarily unavailable.
* A file upload or media generation job fails after waiting.
* Usage, billing, or team settings do not refresh immediately.
* Login, verification, or invite flows do not complete.
* Search or research returns fewer results than expected.
## What to try first
Refresh the page after saving work that you do not want to lose.
Shorten the prompt, use fewer attachments, or split the task into smaller steps.
If the selected model or tool is unavailable, choose another model or try **Manual** tool mode.
Review your API usage budget, plan, team role, and plugin status.
Some external services, connected apps, or model capabilities may recover after a short delay.
## Feature-specific guidance
| Feature | What to try |
| ------------------------- | -------------------------------------------------------------------------------------------------------------------------------- |
| Chat or model replies | Switch models, shorten the request, or start a new thread if the context is full. |
| Web search | Ask for search explicitly, use **Manual > Web search**, and open citations before relying on the result. |
| Deep Research | Open the Canvas report, check whether it is partial or waiting for inline plugin approval, and narrow the scope before retrying. |
| Finance | Refresh the page, use a precise company symbol, check the tool mode, and verify cited sources. |
| Plugins | Refresh **Settings > Plugins**, reconnect if authorization changed, and check your role in the connected app. |
| Image or video generation | Confirm the selected model supports the request and check your usage budget. |
| Memory | If a save, approval, or forget says memory is temporarily unavailable, wait a moment and retry. Contact support if it continues. |
| Files | Try a supported file type, reduce file size, or attach fewer files. |
| Usage or billing | Refresh after a short delay and check whether you are a team owner or member. |
| Login or invites | Use the latest code or invite email and confirm the account email. |
## When to contact support
Contact support if the issue continues after retrying or affects important work.
Include:
* The page or feature you were using.
* The approximate date, time, and timezone.
* The model, tool, plugin, or file type involved.
* The visible error text or a redacted screenshot.
* Whether the issue happens again after refresh.
Do not send passwords, verification codes, private keys, raw connected-app credentials, or sensitive files in a support request.