Mistral OCR

Model

OCR model API that turns PDFs, Office files and images into markdown

Price: Free tier, then paid
Access: API key

About

Mistral's OCR model, served through the Mistral API, reads PDFs, Office files and images and returns markdown per page with tables, bounding boxes and confidence scores. Annotations extract JSON in your schema. Files are limited to 50 MB and 1,000 pages.

What you can do with it

Convert scanned or digital PDFs into markdown with tables and images kept in place
Extract fields from documents into JSON with a schema you define
Batch-process large document sets at half price through the Batch API

Get started

Create an API key in Mistral Studio
Pass a document URL or upload the file first

Example

curl https://api.mistral.ai/v1/ocr \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $MISTRAL_API_KEY" \
  -d '{
    "model": "mistral-ocr-4-1",
    "document": {
      "type": "document_url",
      "document_url": "https://arxiv.org/pdf/2201.04234"
    },
    "table_format": "html"
  }'

Details

Hosting: Hosted service, Self-hosted
Available in: Worldwide
Official SDKs: Python, JavaScript/TypeScript
MCP server: None