Nanonets for AI agents
Nanonets is an intelligent document processing platform that trains models to extract structured data from unstructured documents like invoices and forms. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.
What agents do in Nanonets.
Check a model's extraction results
Pull prediction results from a trained model to review extracted fields, read-only.
Add training images to a model
Upload new sample documents to a model to improve its extraction accuracy on future documents.
Approve before deleting a model
Require approval before deleting an OCR model, since it removes its training data and configuration.
11 Nanonets actions, graded by risk.
Every Nanonets action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
6Look things up. Allowed by default.
- nanonets.get_workflowsGet Workflows
- nanonets.get_all_modelsGet All Models
- nanonets.list_documentsList Workflow Documents
- nanonets.get_model_detailsGet OCR Model Details
- nanonets.get_training_imagesGet OCR Training Images
- nanonets.get_all_prediction_filesGet All Prediction Files
write
4Create and change things. Allow, or ask the user first.
- nanonets.create_modelCreate Model
- nanonets.update_modelUpdate Model AI Guidelines
- nanonets.upload_training_images_by_urlUpload Training Images by URL
- nanonets.upload_training_images_by_fileUpload Training Images by File
destructive
1Delete, cancel or archive. Ask first, or deny outright.
- nanonets.delete_modelDelete OCR Model
Nanonets in three steps.
- 01Your users connect NanonetsThey add their Nanonets api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, writes like “create Model” can wait for the user, and “delete OCR Model” can be denied outright.
- 03Any agent can actYour agent calls Nanonets through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('nanonets', { read: 'allow', write: 'ask', // create_model destructive: 'deny', // delete_model }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Nanonets connects.
Users add their Nanonets api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
The same Nanonets connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →
- AUTH
- API key
- CREDENTIALS
- Per-tenant encrypted vault
- MODEL SEES
- Results only, never credentials
- AUDIT LOG
- Every call, on every plan
Use Nanonets from any agent.
Nanonets and Arc0, answered.
Can I use Nanonets with Claude, ChatGPT or Cursor?
Yes. Connect Nanonets to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Nanonets actions you allow.
How do users connect Nanonets?
Users add their Nanonets api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
Which Nanonets actions can my agent take?
11 in total: 6 read, 4 write and 1 destructive, such as “create Model”. Your policies decide which of them each agent may call.
Can I stop my agent from deleting things in Nanonets?
Yes. Actions like “delete OCR Model” are graded destructive. Set destructive actions to deny, or to ask so the user approves each one, and blocked calls still show up on the audit log.
Can my own backend call Nanonets too?
Yes. The same Nanonets connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Nanonets into your agent.
Your users connect Nanonets once, under your brand. Your agent gets 11 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan