Available on all Cloud plans; Self-Hosted requires Business or higher.
How to Open
Click New chat in the space sidebar to start chatting with Cuppy. To start in a specific project, click + beside its name in the sidebar. You can also pressCtrl + J (Command + J on Mac) to start a new chat in the current project.
After opening a resource such as a table, click the floating 
View resources beside the chat
Click the folder button, Show resources, beside the project name to open the resource panel to the right of the conversation. The resource list has three searchable tabs:- Resources: Shows Recently opened, Starred, and All resources. You can arrange all resources By type or By folder.
- Artifacts: Your artifacts in this project.
- Outputs: Outputs from the current chat, newest first.
What AI Can Reference
AI prioritizes information from the current page:- Current table and view: The current table, current view, and the filtered or sorted results in that view.
- Selected data: Rows, columns, a single cell, or a cell range selected in Grid view become key context for AI. To pin specific data to the chat, choose Add to Chat from the context menu of a Grid view selection, by right-clicking a card or event in Gallery, Kanban, and Calendar views, or from the … menu of an expanded record. The data appears as a chip in the input box. Click the chip to preview the records it contains.
- Uploaded files: Paste or drop PDFs, Excel files, Word documents, images, and other files into the input box. They appear as attachment chips first, then become available to AI after the message is sent. A long block of pasted plain text becomes a Markdown attachment, which you can preview from its chip.
- More nodes with
@: Type@in the input box to add tables, views, apps, automations, or folders from the current project’s tree. - Other projects: Name the project in your request, for example “compare this table with the orders table in the Sales project”, and AI looks it up among the projects you have access to and works there. Anything you do not point elsewhere happens in the current project.
Input Controls
- Voice input: Click the microphone button in the input box and allow microphone access to start recording. While recording, choose Finish voice input or Discard voice input. After you choose finish, Teable transcribes the recording into the input box.
- Model: Use the model menu to choose the model tier for the conversation. The menu lists Ultra, Smart, Standard, and Lite from the most capable down, each showing the model behind it. In Cloud, the menu also marks each tier with its credit multiplier against the default tier, for example 1.0× for the default tier and 2.5× for a higher one. A tier that consumes no platform credit shows no multiplier. Switching tiers during a conversation asks you to confirm once; the conversation is kept, and the new model takes over from your next message. The selected tier and Intelligence level are saved with the conversation and apply again when you reopen it; if the selected tier is no longer offered, the conversation continues on the default tier.
- Intelligence: Choose the Intelligence level separately in the model menu. It controls thinking depth; higher levels produce more thorough reasoning.
- My secrets: Open + → More → My secrets to store API keys or other credentials in Settings → Integrations. The values are yours, AI Chat can read them as environment variables at runtime, and you can grant them to individual apps and automations, so the plaintext never has to go into your prompt. See Credentials and Integrations.
- Authorization cards: When AI needs a third-party account or a secret, it pins a card to the conversation and waits for you before the turn goes on. When Cuppy needs the account to do the task itself, click Connect to run an OAuth flow; when an app or automation needs the credential at run time, choose Connect my account or grant a credential you already have. Click Skip to withhold it; AI continues with the parts it can complete without it. See Credentials and Integrations.
- Scraper: Open the + menu, choose Scraper, then pick a platform and the data type to scrape, such as Posts or Comments, or choose Any web page to paste a link directly. Teable prefills the instruction in the chat input, and you fill in a link, keyword, or name after What to scrape. Each scrape shows in the chat as a Scrape data row with the source and the credits it used. For a platform the menu does not list, describe in plain language which site and what content you want. For examples of this feature in use, read website data collection examples. For practical use cases, see finding and managing influencers.
-
Skills: Open the + menu and choose Skills to import, enable, disable, or try skills in the current chat. When you import one, choose who can use it: Personal keeps it to yourself, Project gives it to collaborators in this project, and Space gives it to everyone in this space. Type
/in the input box to choose an enabled skill. When AI Chat presents a.skillfile, use Install on the file card to add it. To connect or migrate an external system, see Connect & Migrate Everything. - Context usage: The ring next to the input box shows how much of the model’s context window the conversation currently uses. Click it for the percentage and token counts. When the context fills up, Teable compacts the conversation; start a new chat if you want a clean context instead.
- Manage files: Open + → More → Manage files to view files in the current chat sandbox. You can preview supported files, download files, or delete files and folders you no longer need. A conversation must exist before its files can be managed.
- Message queue: Sending another message while AI is working does not interrupt it. The message waits in a queue above the input box and goes out when the current run ends. Queued items keep the same table, view, attachment, and selection chips as the chat input; Remove drops one, … → Edit message puts it back in the input box, and a text-only item also offers Steer to hand its text to the running turn instead of waiting. If the turn is parked on a card waiting for you (an authorization card, a question from AI), a text-only message does not queue: the card closes and is recorded as Skipped, and your message goes straight to that turn as its next instruction, so you do not have to deal with the card first. A message with attachments still goes to the queue.
Manage Chats
The space sidebar groups your chats by project, with Pinned chats at the top. Expand a project to switch chats. If the list does not show all chats, click Show N more. In the chat panel opened beside a resource such as a table, you can also click History to view chats in the current project and search by name. The dot next to a chat name marks which chats need you. Waiting for your reply means the turn has stopped at a point that needs you: a question from the agent, an authorization, a table to pick, or a credential request. It will not continue until you answer. Reply failed means the last reply did not finish properly; open it and ask again. Generating… and the unread dot only report progress and new replies, and need nothing from you. Hover over a chat to Pin or Archive it directly. Right-click a chat (long-press on a touch screen) for the full menu: Rename, Pin, Mark as read or Mark as unread, Archive, and Permanent delete. Permanent delete also removes all of its messages and cannot be undone. After archiving, the notice offers Undo to restore the chat right away and View to open your archived chats. When there are unread chats, a Mark all as read button appears to the right of the search box and clears every unread dot in the list. History shows the 100 most recent chats. Pin the ones you use often: a pinned chat leads the list and is exempt from that limit, so it is never pushed out as new chats pile up. You can also drag a chat to reposition it. App Builder chats are named after their app and cannot be renamed or archived.Archived Chats
Archiving takes a chat out of the history list without deleting anything. Open your avatar at the bottom left → Settings → Archived chats to see chats you archived across projects. Restore brings one back to its project’s chat history, and you can delete it here once you are sure you no longer need it.Common Uses
Create Tables and Automations
AI can create or change tables, views, fields, records, and automations, for example:- Add priority, assignee, due date, and status fields to the task table, then create common views.
- Change the selected records to Completed.
- Create a new table from this Excel file and check whether the field types are appropriate.
- When a new order is created, notify the owner and include the order number, customer name, and amount.
1
Plan first
Describe the business goal, such as building a customer follow-up system, and ask AI to outline the required tables, fields, views, and automations.
2
Confirm the structure
Review the proposed tables, fields, field types, and relationships. Add any missing rules or workflow details.
3
Create next
Ask AI to create the confirmed tables and views, then continue with apps or automations.
Analyze Data
Ask questions about data in the current table, for example:- Count completed tasks this month by assignee.
- Find the customers with the highest sales in the past 30 days.
- Analyze unusual records in this view.
@ to add the related tables or views to the conversation, or use Add to Chat for specific records.
Create Charts and Reports
Ask AI to turn queried data into a chart, a written report, or a small interactive page:- Generate a weekly sales trend chart.
- Turn this query result into a pie chart.
- Create a summary report from these records.
An artifact holds the data AI put into it when it was created, so it does not refresh on its own. For a page that keeps reading live table data, or that collects input and stores state, create an app with App Builder.
Generate Images
Have the AI produce images directly — product concepts, article artwork, poster drafts:- Draw a scene showing this product in use.
- Generate a cover image from the description in this record.
- Take this image and give me three more in winter colours.
AI Chat, the App Builder, and routines can all generate images; chatting with Cuppy in an IM cannot.
Read Files
After uploading files, AI can answer questions using both file content and table data:- Summarize this PDF and connect it with the current customer record.
- Read this Excel file and identify the fields that should be imported.
- Generate a product description from this image.
Memory
Memory stores information you explicitly ask Cuppy to remember, for example by saying “Please remember xxx” in the conversation. Memory is scoped by user and Project, so memory saved in another Project will not automatically appear in the current one. To reuse it, ask Cuppy to read that Project’s memory and save it into the current Project:Handle Long Tasks
AI Chat can keep a single conversation environment available for about 24 hours, which is enough for most data analysis, file processing, and app-building tasks. Keep long-running work in the current conversation so you can follow its progress and provide input when needed.1
Follow progress in the current conversation
If an indicator such as Thinking or Working is still pulsing, or progress updates keep coming in, the task is still running. When a reply finishes, its intermediate steps fold under Worked for {time} and only the final answer stays visible; click it to expand the steps. A reply you stopped partway shows You stopped after {time}.
2
Save results at useful checkpoints
For work with multiple stages, ask Cuppy to write results back to Teable, export them, or generate a downloadable file at key checkpoints. Saved work remains available even if the conversation environment is released later.
3
Know when the environment is released
After foreground execution ends, an idle environment is released after about 30 minutes. Temporary files and unsaved intermediate results are not preserved if the environment restarts.
AI Fields
Use for record-by-record work, such as large table processing, batch classification, content processing, or field filling.
Automation
Use for repeatable work after records are created or updated, notifications, or scheduled data processing.
Artifacts
When AI builds a chart, report, dashboard-style page, or small interactive tool, it saves the result as an artifact: a self-contained page that renders outside the conversation. Artifacts appear as a card in the chat and stay available afterwards. Each one is either an interactive HTML page or a Markdown report. Click the card to open the artifact. From the viewer you can:- Fullscreen or Open in new page to give the artifact more room.
- Download the current version as a file.
- Open the version list to look at an earlier version and Restore this version to make it current again.
- Delete the artifact. Its cards in the conversation turn into a deleted placeholder, so delete only when you no longer need the result.
Share an Artifact
Artifacts are private to the person who created them, so use a share link to let anyone else open one.1
Turn on sharing
Click Share in the viewer and enable Enable sharing. Teable generates the link.
2
Choose who can view
Set Who can view to Anyone with the link, or to Space members so only signed-in members of the space can open it.
3
Add a password (optional)
Under the advanced settings, turn on Password protection. Teable generates a password automatically, and viewers must enter it before the artifact loads. Click Copy link and password to copy both together.
4
Send the link
Use Copy link to copy it. If the link has spread too far, Reset link issues a new one and stops the old link from working.
Find Artifacts Later
Open your avatar at the bottom left → Settings → Manage artifacts. Under Mine, browse the artifacts you created by project or search by name. Items marked Shared have sharing enabled. Use an item’s … menu to Share, Download, Open in new page, or Delete it. Artifacts that others shared and you opened while signed in are listed under Shared with me, with who shared them and when you last opened them; click one to reopen it in a new tab. If the share is turned off, expires, or is deleted, or its link is reset, the entry drops off the list. Click an item to open the viewer, inspect its content and version history, or adjust sharing. You can also open the list from a chat through + → More → Manage artifacts.
Prompting Tips
- State the goal first: In the first message, describe what to do, the relevant conditions, and the expected output. For example, analyze weekly sales trends over the past 3 months, grouped by region, and return a trend table with a short conclusion.
- Use field names: For example, sort by
Created Time, so AI can identify the target column. - Provide several examples: For fixed-format output, provide 3-5 input -> expected output examples.
- Confirm before execution: For tasks that modify data, create tables, create apps, or enable automations, ask AI to describe the changes before running them.
- Choose the tier by task: A lower tier costs less credit on simple queries, cleanup, and rewriting; move up a tier for complex planning, cross-table analysis, and app building.
- Start a new chat when the topic changes: A new chat is recommended when the topic changes, the conversation becomes too long, or context usage should be reduced.
FAQ
Can my teammates open an artifact I created?
Can my teammates open an artifact I created?
Not by default. Only you can open your own artifacts. Turn on Enable sharing in the viewer and send the link, choosing Space members if the artifact should stay inside the space.
How many images can one request produce?
How many images can one request produce?
Five by default. Ask the AI to work in several rounds when you need more. The number of images a space can have in flight at once is capped as well, and anything beyond it waits in the queue.
Which model generates the images?
Which model generates the images?
The available image models come from the project’s AI configuration: the space’s own models first, then the models Teable provides. When the project has no image model available, the AI does not offer image generation at all.
How are generated images charged?
How are generated images charged?
An image counts against credits only once it is saved; images that fail or are cancelled cost nothing. In Credit usage summary on the billing page, these charges appear under the type AI image generation. When the space has its own models (BYOK), generating with them is not blocked by exhausted credits.
Why does Teable say Cuppy is busy in another conversation?
Why does Teable say Cuppy is busy in another conversation?
Each user can run up to 3 Agents at the same time, including AI Chat and App Builder. If you see
Agent is busy in another conversation. Please wait., another task is still running. Wait for Cuppy to finish the previous task before continuing.How are credits refunded when an AI run fails?
How are credits refunded when an AI run fails?
One AI Chat turn can include multiple processing steps. The conversation and billing page both show the total credits for that turn. Work that Cuppy already completed and saved can still be used. If the final step is interrupted by an unexpected error, credits for that interrupted step are refunded automatically. You can continue working in the same conversation.
What happens if I send another message while AI is working?
What happens if I send another message while AI is working?
It joins the send queue rather than interrupting the run, and Teable sends the whole queue as one message when the run ends. The queue lives on the server, so it survives a refresh and looks the same in every tab or device where that conversation is open. A message handed to the running turn with Steer appears in the conversation labelled Steered conversation.A conversation holds up to 30 queued messages. Past that, the message is not queued and Teable replies
The send queue is full. Wait for Cuppy to catch up.Why does the queue say it is paused?
Why does the queue say it is paused?
Stopping a run also holds the queue, so nothing you queued is sent by a stop you meant for the current answer. The queue header then reads
Queue paused because you interrupted. Click Resume to release it, or just send your next message: starting a new run releases the queue as well.Who can add project skills and space skills?
Who can add project skills and space skills?
Users who can manage the project can add project skills, and space owners and creators can add space skills.
Which files stay in Manage files?
Which files stay in Manage files?
Manage files lets you browse two folders:
uploads for the files you added, and outputs for the files produced in the conversation. Both use their own storage, so they survive the sandbox being released or rebuilt and you can come back later to preview, download, or delete them. The root also lists config and home, which hold tool configuration and state. The agent manages them, so you can’t open, download, or delete them here, but they still count toward storage usage.Anything Cuppy writes outside those four folders is temporary, such as intermediate results from a script. It does not appear in Manage files and is gone after the sandbox restarts. For results you need to keep, ask Cuppy to write them into outputs or back into Teable, or download them while the conversation is open.What happens when I copy a message with chips?
What happens when I copy a message with chips?
When you copy a user message, tables, views, apps, automations, folders, selections, and attachments are kept as marker text. When you paste the text back into AI Chat, Teable restores the markers it can recognize as chips.
How much data does one scrape return, and what does it cost?
How much data does one scrape return, and what does it cost?
For lists such as search results, posts, and comments, the AI fetches a small number of results by default; state a number in your request if you need more. Each scrape uses credits at the source’s actual price. Before it runs, Teable holds the most the scrape could cost, and the scrape cannot run if the space does not have enough credits. A scrape whose maximum cost exceeds 150 credits is refused; ask for fewer results or split it into several scrapes.