MCP Mistral OCR

What is MCP Mistral OCR?

MCP Mistral OCR is an MCP server that provides OCR capabilities using Mistral AI's OCR API. It can process both local files and URLs, supporting images and PDFs.

How to use MCP Mistral OCR?

Install via Smithery (npx -y @smithery/cli install @everaldo/mcp/mistral-crosswalk --client claude), Docker, or local development with UV. Set the MISTRAL_API_KEY and OCR_DIR environment variables. Use the process_local_file or process_url_file tools to extract text from files.

Key features of MCP Mistral OCR

• Process local files (images and PDFs) using Mistral's OCR
• Process files from URLs with explicit file type specification
• Support for JPG, PNG, PDF, and other formats
• Results saved as JSON files with timestamps
• Available as Docker container
• Uses UV package management

Use cases of MCP Mistral OCR

• Extract text from scanned documents stored locally
• Transcribe text from images hosted on the web
• Automate OCR processing in Claude Desktop workflows
• Batch processing of PDF files under 50MB and 1000 pages

FAQ from MCP Mistral OCR

What file types does MCP Mistral OCR support?

It supports images (JPG, JPEG, PNG, GIF, WebP) and PDF documents.

What are the file size and page limits?

Files must be under 50MB and documents under 1000 pages, as enforced by the Mistral API.

How do I provide my Mistral API key?

Set the MISTRAL_API_KEY environment variable when running the server, either in Docker or in your local environment.

Where are OCR results saved?

Results are saved as JSON files in the output directory inside the configured OCR_DIR path.

Can I use MCP Mistral OCR without Docker?

Yes, you can install it locally using UV and the uv pip install . command, then run it as an MCP server.