Integration · Documents

Connect Claude & ChatGPT to Sensible

Ask your AI to pull the fields out of an invoice or certificate with Sensible, check whether an extraction has finished, or export last week's results to a spreadsheet. It can extract from a document you send or from a public URL, read document types and their configurations, and list past extractions with their review status. You get the data without opening the Sensible app.

Connect Sensible free

Free plan, no credit card. Takes about a minute.

Try asking

“Extract the data from the invoice PDF at this link.”
The AI starts an extraction with your invoice document type, then fetches the result and shows the extracted fields once it is complete.
“Which document types do we have set up in Sensible?”
It lists your document types with their IDs and names, and can read the output schema of the one you pick.
“Show me last week's extractions that need review.”
It lists extractions in that date range with the review status NEEDSREVIEW.

What your AI can do

Extract from documents

Extract from a document sent directly, from a public URL, with one specific configuration, or from a PDF that bundles several documents.

Results and exports

Read any extraction with its status and fields, list past extractions by date, status or review status, and export them to CSV or Excel.

Document types and configurations

List document types and read their output schema, list and read extraction configurations, and see usage statistics.

“Export all of yesterday's completed extractions to Excel.”
It finds the extractions and returns a download link for one Excel file that combines them.
“Is extraction 8f2c1a done yet, and what did it find?”
It reads the extraction and shows its status (waiting, processing, complete or failed) and the extracted fields when it is complete.
“How many documents did we extract in September?”
It reads extraction counts and usage statistics for the date range.

13 tools for Sensible

These tools are switched on when you connect. You can switch any of them off, or require your approval before it runs.

export_extractions_csvGet a download link for a CSV file combining one or more extractions, by extraction IDs in `ids`.
export_extractions_excelGet a download link for an Excel file combining one or more extractions, by extraction IDs in `ids`.
extract_documentChanges dataExtract structured data right away from a document sent as base64 text in `document`, using the `document_type` name. Returns the extracted fields in the response; set `content_type` for non-PDF files.
extract_from_urlChanges dataStart extracting structured data from a document at a public `document_url`, using the `document_type` name; Sensible picks the best configuration. Runs asynchronously: use get_extraction with the returned ID to fetch the result.
extract_from_url_with_configChanges dataStart extracting data from a document at `document_url` using one specific `document_type` and `config_name`. Runs asynchronously: fetch the result with get_extraction.
extract_portfolio_from_urlChanges dataStart extracting data from a PDF at `document_url` that bundles several documents, listing the document type names it contains in `types`. Runs asynchronously: fetch the result with get_extraction.
get_configurationGet one extraction configuration by `type_id` and `config_name`, including its published versions.
get_document_typeGet one document type by `type_id`, including its name and output schema.
get_extractionGet one extraction by `document_id`, including its status (WAITING, PROCESSING, COMPLETE or FAILED) and the extracted fields once complete.
get_extraction_statisticsGet extraction counts and usage statistics between `start_date` and `end_date`.
list_configurationsList the extraction configurations for a document type (`type_id`). Use to find a `config_name` when you want to extract with one specific configuration.
list_document_typesList the document types set up in your Sensible account (for example invoices or insurance certificates), with their IDs and names. Use to find the `document_type` name before extracting.
list_extractionsList past extractions, filtered by `start_date`/`end_date`, `statuses`, `review_statuses` (APPROVED, NEEDSREVIEW, REJECTED), `document_type_ids` or coverage score. Pages with `continuation_token`.

Set up in three steps

1
Pick the app
Create a free PipMCP account and choose Sensible from the app list.
2
Paste your key
Paste your Sensible API key. In the Sensible app, open your Account page (app.sensible.so/account) and create or copy an API key. Keep it secret.
3
Add the link to your AI
You get a personal MCP link. Add it to Claude, ChatGPT or Cursor:
1. Click your name, then Settings › Connectors › Add custom connector.
2. Paste your link as the Remote MCP server URL.
3. Switch it on from the + menu in a chat.

Questions

What can the AI do in Sensible?

It can extract structured data from a document sent as base64 or from a public URL, with or without a specific configuration, and from a PDF that bundles several documents. It can read extractions and their fields, list past extractions, export them to CSV or Excel, read document types and configurations, and get usage statistics.

Does the AI see my Sensible credentials?

No. Your Sensible API key is encrypted at rest and never shown to the AI. After you save it, it is not shown again, not even to you. The AI only sees the results of the tools it calls.

Can I control what the AI is allowed to do?

Yes. You choose which tools are switched on, so you can start by only reading past extractions. Each new extraction runs in your Sensible account and can require your approval before it runs. Every tool call is logged.

Does it work with ChatGPT?

Yes. In ChatGPT go to Settings › Apps & Connectors › Advanced and turn on Developer mode, then add your PipMCP link. Developer mode needs a paid ChatGPT plan: Plus, Pro, Business or Enterprise. The same link also works in Claude (Settings › Connectors › Add custom connector), Cursor and other MCP clients.

What does it cost?

PipMCP has a free plan with no credit card required. Paid plans bill per completed task. You also need your own Sensible account.

Why does an extraction from a URL not return data right away?

Extractions from a URL run asynchronously in Sensible. The AI starts the extraction, gets an ID back, and then fetches the result with that ID once the status is complete. A document sent directly as base64 returns its fields in the same response.