mistral_ocr

MCP tool from Mistral MCP by Swih

Document AI — extract text, bbox, and JSON annotations from PDFs/images. Pass includeBlocks: true for OCR 4 paragraph-level blocks (text/title/table/image/equation/... with bounding boxes).

How to use it

mistral_ocr is exposed by the Mistral MCP MCP server. Add the server to your MCP client (Claude Desktop, Cursor, Windsurf and others), and the mistral_ocr tool becomes available to the model automatically. See the full listing for setup details and every tool this server provides.

Install Mistral MCP

$claude mcp add mistral -- npx -y mistral-mcp@latest

FULL MISTRAL MCP LISTING

Other tools in Mistral MCP (12)

codestral_fim

Fill-in-the-middle code completion (Codestral model).

connectors_call_tool

Invoke a connector's tool — real MCP CallToolResult passthrough.

connectors_get

Fetch one connector's public metadata (never credentials).

connectors_list

Discover Mistral Connectors (MCP/HTTP integrations) visible to the caller.

connectors_list_tools

List the MCP tools a connector exposes, with their input schema.

mistral_chat

Chat completion. Supports all Mistral models, responseformat, reasoningeffort for Magistral.

mistral_vision

Multimodal chat with images (URL or base64).

process_document

Single-call macro-tool: OCR → classify (kind=auto) → typed extraction → validation → cache. Kinds: contract / invoice / iddocument / generic. Returns a discriminated union. PII-safe cache (iddocument auto-bypass). Configurable minOcrConfidence.

voxtral_transcribe

Audio → text. Pass diarize: true for speaker separation.

workflow_execute

Start a Mistral Workflow (Temporal-backed durable execution).

workflow_interact

Signal / query a running workflow. Used for human-in-the-loop checkpoints.

workflow_status

Poll a running workflow — returns RUNNING \