Mistral MCP — Document Extraction
mistral-mcp
Servidor MCP que expone las capacidades de Mistral AI a cualquier cliente MCP - Claude Code, Cursor, Zed, Windsurf, Claude Desktop.
Versión francesa: README.fr.md
Por qué
Mistral cuenta con modelos potentes para francés, código, OCR, moderación, audio y flujos de trabajo tipo agente, pero la mayoría de los IDE compatibles con MCP utilizan Anthropic u OpenAI por defecto. mistral-mcp proporciona a esas capacidades de Mistral una superficie MCP limpia para que puedas dirigir la subtarea correcta al modelo adecuado sin tener que reconstruir tu bucle de agente.
El objetivo de este repositorio no es ser "otro envoltorio ligero más". Aspira a ser un servidor MCP robusto y mantenible con esquemas explícitos, resultados predecibles, flexibilidad de transporte y una buena cobertura de pruebas.
Related MCP server: MCP Server TypeScript
Superficie actual (v0.4.0)
Herramientas (22)
Generación principal:
mistral_chatmistral_chat_streammistral_embedmistral_tool_callcodestral_fim
Visión y audio:
mistral_visionmistral_ocrvoxtral_transcribevoxtral_speak
Agentes y clasificadores:
mistral_agentmistral_moderatemistral_classify
Archivos y procesamiento por lotes:
files_uploadfiles_listfiles_getfiles_deletefiles_signed_urlbatch_createbatch_listbatch_getbatch_cancel
Utilidad nativa de MCP:
mcp_sample- delega la generación al modelo cliente mediante muestreo MCP
Recursos (2)
mistral://models- alias aceptados y catálogo de modelos en vivomistral://voices- catálogo de voces en vivo para Voxtral TTS
Prompts (6)
Prompts seleccionados en francés:
french_invoice_reminderfrench_meeting_minutesfrench_email_replyfrench_commit_messagefrench_legal_summary
Prompt seleccionado en inglés:
codestral_review
Los argumentos de enumeración de los prompts están envueltos con completable(), por lo que los clientes MCP pueden solicitar la finalización de argumentos de prompt mediante completion/complete.
Aspectos destacados
API
McpServerde alto nivel coninputSchema,outputSchemay anotaciones en cada herramientaSoporte de transporte dual: stdio por defecto, HTTP transmitible para despliegues remotos
Resultados estructurados en todas partes:
structuredContentmás alternativa de textoSoporte de muestreo MCP a través de
mcp_sampleSoporte de finalización de prompts para argumentos de tipo enumeración
Recursos y prompts registrados junto con las herramientas, no añadidos posteriormente
Reintento/retroceso y tiempo de espera de solicitud en el cliente del SDK de Mistral
Transporte
Stdio
Modo predeterminado. Es lo que utilizan Claude Code y la mayoría de los clientes MCP locales.
node dist/index.jsHTTP transmitible
Habilitar con --http o MCP_TRANSPORT=http.
MCP_TRANSPORT=http node dist/index.jsVariables de entorno relevantes:
MCP_HTTP_HOST- predeterminado127.0.0.1MCP_HTTP_PORT- predeterminado3333MCP_HTTP_PATH- predeterminado/mcpMCP_HTTP_TOKEN- token de portador opcionalMCP_HTTP_ALLOWED_ORIGINS- lista de permitidos opcional separada por comasMCP_HTTP_STATELESS=1- modo de sesión sin estado
/healthz es intencionalmente público y no toca el servidor MCP.
Instalación
git clone https://github.com/Swih/mistral-mcp.git
cd mistral-mcp
npm install
npm run buildEstablece tu clave de API:
export MISTRAL_API_KEY=your_key_hereO utiliza .env en la raíz del repositorio. Nunca lo subas al control de versiones.
Uso en Claude Code
claude mcp add mistral -- node /absolute/path/to/mistral-mcp/dist/index.jsEjemplo de prompt:
Usa
mistral_ocren este PDF, luego ejecutafrench_meeting_minutesen el texto extraído.
Desarrollo
npm run dev
npm run build
npm run lint
npm test
npm run inspectorEstrategia de pruebas
El conjunto contiene actualmente 148 pruebas en 4 capas:
Pruebas unitarias para herramientas, recursos, prompts, transporte, audio, agentes, archivos, procesamiento por lotes y muestreo
Pruebas de contrato para metadatos de herramientas y garantías orientadas a MCP
Pruebas de API en vivo contra la API real de Mistral cuando se establece
MISTRAL_API_KEYPruebas de extremo a extremo de Stdio contra el servidor compilado
Sin MISTRAL_API_KEY, el valor predeterminado local es 139 aprobadas más 9 pruebas en vivo/stdio bloqueadas.
Diseño del proyecto
mistral-mcp/
|-- src/
| |-- index.ts
| |-- transport.ts
| |-- tools.ts
| |-- tools-fn.ts
| |-- tools-vision.ts
| |-- tools-audio.ts
| |-- tools-agents.ts
| |-- tools-files.ts
| |-- tools-batch.ts
| |-- tools-sampling.ts
| |-- resources.ts
| `-- prompts.ts
|-- test/
|-- examples/
|-- .github/workflows/ci.yml
|-- package.json
`-- tsconfig.test.jsonEstado
v0.4.0 — publicado. Consulta CHANGELOG.md para ver la diferencia completa respecto a v0.3.0:
ayudantes compartidos, catálogos de modelos + voces en vivo, pruebas de contrato
visión + OCR
transcripción de audio + voz
agentes + moderación + clasificación
APIs de archivos + procesamiento por lotes
transporte HTTP transmitible + muestreo MCP
5 prompts seleccionados en francés + 1 prompt en inglés + finalización de argumentos de prompt
Ejemplos
Los scripts ejecutables se encuentran en examples/. Consulta examples/README.md.
Licencia
MIT Copyright Dayan Decamp
Available Tools
6 toolscodestral_fimCodestral fill-in-the-middle completionARead-only
Fill-in-the-middle code completion with Codestral.
Given prompt (code preceding the cursor) and suffix (code after the cursor),
Codestral writes the middle. Use for editor autocomplete scenarios, code-patching
agents, or structured refactors where you know the target boundaries.
Default stop tokens: [] — let the model decide. Override with stop if needed.
Inputs: prompt (required), suffix (required), seed, stop, +4 more. Returns: text, model, usage, finish_reason
mistral_chatMistral chat completionARead-only
Generate a chat completion using a Mistral model.
When to use:
Drafting French (or any European-language) content where Mistral shines.
Codestral for code-specific generation/review.
Ministral for cheap / low-latency classification.
Returns structured content with the assistant text and token usage. Does NOT stream — use mistral_chat_stream for long outputs with progress updates.
Inputs: messages (required), seed, model, top_p, +4 more. Returns: text, model, usage, finish_reason, +1 more
mistral_ocrMistral OCR (document to markdown)ARead-onlyIdempotent
Run Mistral OCR on a PDF or image, returning structured markdown per page.
Input document is one of:
{ type: "document_url", documentUrl: "https://...pdf" }
{ type: "image_url", imageUrl: "https://..." | "data:image/..." }
{ type: "file", fileId: "" }
Options:
pages: array of 0-indexed page numbers or string like "0-5,7".tableFormat: 'markdown' (default) or 'html'.extractHeader/extractFooter: include page header/footer when present.includeImageBase64: embed extracted image bytes as base64 in the response.document_annotation_format: JSON schema for whole-document structured extraction.bbox_annotation_format: JSON schema for extracted image / bbox annotations.confidence_scores_granularity: 'page', 'word', or 'block'. 'block' adds per-block content/type confidence underpages[].blocks[].confidence_scoresand requires OCR 4.1 or newer.includeBlocks: return paragraph-level blocks (bounding box + type) in reading order — titles, lists, tables, images, equations, captions, code, references, aside text, header, footer, signature. Requires OCR 4 (mistral-ocr-4-0) or newer; older models accept the flag but return an emptyblocksarray.
Returns pages[].markdown plus optional pages[].hyperlinks, header, footer,
images bounding boxes, blocks, annotations, confidence scores, and dimensions.
Inputs: document (required), model, pages, imageLimit, +10 more. Returns: model, pages, usage, annotations, +2 more
mistral_visionMistral multimodal chat (vision)ARead-only
Chat completion with multimodal input: text + image_url parts.
Requires a vision-capable model. Accepted:
pixtral-large-latest
pixtral-12b-latest
mistral-large-latest
mistral-medium-latest
mistral-small-latest
Each message's content is either a plain string (pure text) or an array of
parts { type: 'text', text } / { type: 'image_url', imageUrl }. The image URL
can be an https URL or a data: URI base64 payload.
Returns the assistant text + token usage. For non-visual requests, prefer mistral_chat.
Inputs: messages (required), seed, model, top_p, +2 more. Returns: text, model, usage, finish_reason
process_documentProcess a business document end-to-endARead-onlyIdempotent
Single-call pipeline: provided text/Markdown or Mistral OCR → classify (if kind=auto) → typed extraction → validation. source.type=text skips OCR and Files uploads. Text with kind=generic makes no API calls; classification and typed extraction use Mistral chat. Results expose extraction_source. Provided text has null ocr_confidence and page_count; ocr_text contains the supplied text unchanged. Replaces the manual chain of mistral_ocr + mistral_chat + JSON parsing.
Kinds: contract | invoice | id_document | generic. Use kind=auto to let the server classify.
Returns a discriminated union — switch on kind to access typed fields.
Validation checks schema and, for OCR sources, OCR confidence; not factual or accounting accuracy.
Typed extraction rejects text longer than 60000 characters rather than truncating it.
Cache keys include source, kind, page limit, endpoint, models and pipeline version. Override location with MISTRAL_MCP_CACHE_DIR. Override mode with options.cache. Default cache mode is 'read_write' EXCEPT for kind=id_document (auto-bypass to avoid persisting PII). Set options.cache='read_write' explicitly to opt in for id documents.
options.maxPages and options.minOcrConfidence apply only to OCR sources. The confidence floor defaults to 0.3. Below the floor the
tool returns isError. Missing or partial confidence scores also return isError;
use mistral_ocr directly if you need raw OCR without a confidence guarantee.
0.3 is a conservative starting point, not a measured one: calibrate it for your
corpus with npm run eval:docs.
Inputs: source (required), kind, options. Returns: dob, kind, name, total, +22 more
voxtral_transcribeVoxtral speech-to-textARead-onlyIdempotent
Transcribe an audio file to text using Mistral Voxtral.
Accepted models:
voxtral-mini-latest
voxtral-small-latest
Audio source is one of:
{ type: "file_url", fileUrl: "https://..." } (public URL)
{ type: "file", fileId: "" }
Options:
language: ISO-639-1 hint (e.g. 'fr', 'en'). Boosts accuracy when known.temperature: sampling temperature.diarize: return per-speaker segments (default false).timestampGranularities: ['segment'] to return per-segment timestamps.contextBias: list of phrases/terms that should bias the decoder.
Returns plain text, detected language, optional segments[], and token usage.
Inputs: audio (required), model, diarize, language, +3 more. Returns: text, model, usage, language, +1 more
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 6 tools
Each tool targets a distinct modality/action: code FIM, text chat, OCR, vision chat, document pipeline, and audio transcription. There is minor overlap between mistral_chat and mistral_vision (both chat) and between process_document and mistral_ocr, but the descriptions explicitly steer selection, so confusion is limited.
Names mix provider-prefixed conventions (codestral_fim, mistral_chat, mistral_ocr, mistral_vision, voxtral_transcribe) with a plain action name (process_document). The pattern is somewhat readable but not a predictable verb_noun scheme and the provider prefixes fragment consistency.
Six tools is well-scoped for a multimodal Mistral wrapper, with each tool covering a distinct capability rather than redundancy. Slightly broad for a server named 'Document Extraction' since it also includes code completion and chat, but not excessive.
Covers OCR, vision, chat, transcription, and a combined pipeline, but several tools reference a Files API (fileId) that is not exposed as a tool, creating a dead-end for uploading files. No model-listing or file-management operations leave workflows partially blocked.
Maintenance
Related MCP Connectors
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
MCP server for progressive tool usage at any scale (see https://klavis.ai)
214 tools for AI agents over one MCP endpoint: image, video, audio, 3D, data, storage, mail, GPUs.
Talk to your public-facing AI from any MCP client — Claude, ChatGPT, Cursor, Cline, Windsurf.
Related MCP Servers
- -licenseNot gradedqualityDmaintenanceA TypeScript implementation of a Model Context Protocol server and client that enables interaction with language models (specifically Mistral running on Ollama).-
- -licenseNot gradedqualityNot gradedmaintenanceA production-ready TypeScript MCP server providing basic tools (add, echo, timestamp), resources (server info, greetings, data access), and prompt templates (analyze, code-review, summarize). Serves as a foundation for building custom MCP servers with extensible architecture.547 npm-
- AlicenseBqualityDmaintenanceA TypeScript-based MCP server that provides tools to interact with local Codex and Gemini CLIs via stdio transport. It enables users to execute prompts through the ask_codex and ask_gemini tools, supporting custom models and timeout configurations.111 npm6MIT
- FlicenseAqualityDmaintenanceA TypeScript MCP server template with Zod validation, dual transport (stdio/HTTP), and modular architecture for building MCP-compatible tools, resources, and prompts.11-