Nanonets MCP server
17 operations0 webhooks
AI document processing with Nanonets - extract fields from images and URLs with OCR models, train models, review and correct predicted files, approve or delete them, and query connected data stores.
Try Nanonets with your agent
Connect Nanonets, then paste this into Claude, ChatGPT or Cursor. For ideas from other apps, see the use cases.
Using Corsair, show me what you can do in Nanonets on my account, then run one read-only example and show me the result.
Nanonets tools and triggers
Tools are what your agent can do in Nanonets through Corsair's MCP server. Triggers are the Nanonets events a routine can start from.
Files
- ApproveWrite
Approve a validated file from an approval stage, which can trigger exports
files.approve
- AssignWrite
Assign files to a team member by email for review
files.assign
- DeleteDestructive
Permanently delete a file and its predictions from a model
files.delete
- Retry ExportsWrite
Re-run the export integrations for the given files
files.retryExports
- Retry PredictionsWrite
Re-run OCR prediction for the given files
files.retryPredictions
- UnapproveWrite
Reopen a previously approved file; it cannot be rejected here
files.unapprove
- Update FieldsWrite
Replace the field set of one page of a file: unchanged boxes stay, new boxes need is_new, and any omitted field is deleted
files.updateFields
Predictions
- Create From FileWrite
Upload a base64 PNG or JPEG to an OCR model and extract fields synchronously or, with async, as a pollable request
predictions.createFromFile
- Create From UrlsWrite
Extract fields from public image URLs with an OCR model, synchronously or as a pollable async request
predictions.createFromUrls
- Get File By File IdRead-only
Fetch every page of a prediction file by request file ID, listed as moderated and unmoderated; the poll target for async requests
predictions.getFileByFileId
- Get File By Page IdRead-only
Fetch one prediction page with its boxes and raw OCR output
predictions.getFileByPageId
- List FilesRead-only
List a model prediction files in a day-since-epoch window, split into moderated and unmoderated
predictions.listFiles
Models
- TrainWrite
Train or retrain a model from its uploaded images; the model state moves through the training queue
models.train
- Update GuidelinesWrite
Add or update the natural-language field and table extraction guidelines of an Instant Learning model
models.updateGuidelines
- Upload Training UrlsWrite
Add training images to a model by URL, with optional box annotations, before (re)training
models.uploadTrainingUrls
Integrations
- Execute QueryDestructive
Run arbitrary SQL against a connected store; documented examples include UPDATE, which cannot be undone
integrations.executeQuery
- ListRead-only
List external integrations with their IDs, types and masked connection info
integrations.list
Connect Nanonets to Claude, ChatGPT or Cursor
- 1
Connect Nanonets in Corsair
Add your Nanonets API key once in Corsair. Corsair stores the credential encrypted and makes the Nanonets API calls for your agent, so it never goes into a prompt or a config file. Connect Nanonets
If your link lets the agent add apps, you can skip this: it gives you a sign-in link the first time it needs Nanonets. On a ChatGPT personal plan, connect Nanonets here first.
- 2
Add Corsair's MCP server to your agent
Every client uses the same address:
https://mcp.corsair.dev/mcpOpen Customize → Connectors → Add custom connector, paste the address, then select Connect.Add a custom MCP server with the address and choose OAuth. Personal plans get read tools only.Add to CursorOr add this to~/.cursor/mcp.json. Cursor opens Corsair to sign in the first time it connects.{ "mcpServers": { "corsair": { "url": "https://mcp.corsair.dev/mcp" } } }Run this, then run/mcpin Claude Code to sign in.claude mcp add --transport http corsair https://mcp.corsair.dev/mcpCodex, Gemini CLI, VS Code, Zed and other clients with OAuth sign-in work the same way. Clients without it, such as GitHub Copilot's coding agent, use a link's own address and an API key. Setup for every client
- 3
Sign in and choose what needs your approval
Your client opens Corsair to sign in. Pick a link: the apps your agent can use and a mode that decides what it may do without asking. In the default Cautious mode, Nanonets's read and write operations run right away and its 2 destructive operations wait for your approval. Strict mode asks before every write and refuses destructive operations, and Read mode allows reads only.
FAQ
Install corsair and @corsair-dev/nanonets, add the Nanonets plugin, and connect with an API key. Then call methods like corsair.nanonets.api.files.approve() from your app. Corsair keeps each tenant isolated.
Yes. Nanonets requires an API key from your Nanonets account or developer dashboard. Create credentials with access that matches what you automate, then paste them into Corsair or connect through managed auth in Hub.
Yes. One Corsair client can hit Nanonets, GitHub, Notion, Stripe, and the rest. Trigger Nanonets from a webhook, push updates to Slack, or chain actions across plugins. Same tenant context across integrations, plain TypeScript.
Teams use Corsair for AI document processing with Nanonets - extract fields from images and URLs with OCR models, train models, review and correct predicted files, approve or delete them, and query connected data stores. Common builds include sync jobs, admin tooling, and agents that call Nanonets APIs from your backend.
The SDK is open source. Self-host and you pay Corsair nothing for the integration layer. Hub bills for hosted auth and execution, not per Nanonets API call. Hub has a generous free tier. For hobby projects you can probably run the whole thing on free. Paid plans start to matter with more tenants or heavier traffic.
We maintain @corsair-dev/nanonets in the open repo. When Nanonets ships API changes, upgrade the package. Your call sites stay typed. You should not need to rewrite the integration layer each time.
No. Corsair does not read your Nanonets data or hold individual access credentials. Credentials stay in your database. API calls run on your instance. We do not store Nanonets data on your behalf.
Yes. Add https://mcp.corsair.dev/mcp as a custom connector or MCP server in Claude, ChatGPT, Cursor, VS Code, Claude Code, Codex or any client that supports remote MCP servers with OAuth sign-in, then sign in to Corsair. On ChatGPT personal plans, custom servers currently get read tools only.
Yes. Put Nanonets on a link in Strict mode and every write waits for your approval, while destructive actions are refused. The default Cautious mode asks only before the 2 destructive Nanonets operations. You can change a link's mode at hub.corsair.dev/me.