DeepSeek Tools MCP
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@DeepSeek Tools MCP look up the latest DeepSeek API pricing and fetch the first result"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
deepseek-tools-mcp
Unofficial MCP server that exposes DeepSeek's standalone tools APIs as MCP tools: Web Search, Fetch, OCR, ASR and TTS.
非官方项目,与 DeepSeek 无隶属关系。DeepSeek 及其 API 的条款与定价以官方文档为准。 This project is not affiliated with DeepSeek. Use it under DeepSeek's own terms and pricing.
Why this exists
DeepSeek opened its standalone tools API in 2026 and added OCR / ASR / TTS shortly after. Those
are plain REST endpoints, which makes them a natural fit for MCP — but most existing MCP wrappers
only cover Search and Fetch, and the in-conversation server tools (deepseek_web_search,
deepseek_read_url) cannot be reused outside a model round-trip.
This server wraps the standalone APIs so any MCP client can call them directly:
MCP tool | DeepSeek endpoint | Notes |
|
| Query operators, |
|
| Page to Markdown, offset paging for long pages |
|
| PDF / Word / PowerPoint / images, per-page text |
|
| WAV (16-bit PCM) or Ogg Opus, ≤8 MiB / ≤120 s |
|
| wav / pcm, written to a local file |
It deliberately does not wrap deepseek_web_search or deepseek_read_url: those are
server-side tools that require a model round-trip, and the standalone Search/Fetch APIs cover the
same ground more cheaply.
Related MCP server: mcp-deepseek-search
Requirements
Node.js >= 20
A DeepSeek API key with access to the tools API
Quick start
DEEPSEEK_API_KEY=sk-... npx -y deepseek-tools-mcpFrom source:
git clone https://github.com/wuzhechen/deepseek-tools-mcp.git
cd deepseek-tools-mcp
npm install
npm run build
DEEPSEEK_API_KEY=sk-... node dist/src/index.jsThe server speaks MCP over stdio; stdout is the protocol channel, logs go to stderr.
Client configuration
Claude Code
claude mcp add deepseek-tools \
--env DEEPSEEK_API_KEY=sk-... \
-- npx -y deepseek-tools-mcpCC Switch (and dsh)
Add a new stdio MCP server:
{
"type": "stdio",
"command": "npx",
"args": ["-y", "deepseek-tools-mcp"],
"env": { "DEEPSEEK_API_KEY": "sk-..." }
}CC Switch writes MCP entries into each app it manages — for DeepSeek Harness that means
~/.dsh/profiles/<profile>/cordis.patch.yml as a @deepseek-ai/dsh-mcp-client entry, so the
same tools become available to Claude Code and dsh.
Generic MCP config
See examples/mcp.json.
Environment variables
Variable | Default | Description |
| — (required) | DeepSeek API key. Read from the server process environment only. |
|
| Override for proxies or private deployments. |
|
| Cap on characters returned per call (search / fetch / ocr). Minimum 1000. |
| unset (any path) | Optional |
Tool reference
deepseek_search
Input | Type | Description |
| string, 1–128 chars | Supports |
| int, 1–50, default 10 | Number of results. |
| string, ≤128 chars | Original user question, improves recall. |
| boolean, default false | Ask for page content too (heavier). |
| string | Optional locale/IP hint. |
Returns numbered results with title, URL, summary/snippet, date and authority.
deepseek_fetch
Input | Type | Description |
| string | http/https only, ≤8192 chars. |
| int, default 0 | Character offset for paging. |
| int, 1000–50000, default 20000 | Chunk size per call. |
Returns URL, the character range and the Markdown content. When the page is longer, the header
includes the next offset.
deepseek_ocr
Input | Type | Description |
| string | PDF, doc/docx, ppt/pptx, png/jpeg/webp/gif/bmp/tiff. Max 100 MiB, 1000 pages. |
|
|
|
Returns page-count and --- page N --- blocks. The API budget is 180 s end-to-end; this tool waits
up to 200 s.
deepseek_asr
Input | Type | Description |
| string | 16-bit PCM WAV or Ogg Opus. Max 8 MiB and 120 s. Stereo uses the left channel. |
Returns text and duration_ms.
deepseek_tts
Input | Type | Description |
| string | Non-empty. Up to 10000 billed characters (Chinese counts as 2) and a 64 KiB request body. |
|
| Female / male voice. |
|
|
|
| boolean, default true | Strip Markdown before synthesis. |
| string | Optional. Defaults to |
Audio is written to a file and the tool returns the path, format, content type and byte size.
Limits and pricing
Limits below are enforced by the API; see the official docs for the current values.
Search / Fetch: per-request pricing (about ¥5 per 1000 successful calls), rate limited per account.
OCR: max 100 MiB, 1000 pages, 180 s total; about ¥1.5 per 1000 pages.
ASR: max 8 MiB, 120 s; priced per audio hour.
TTS: max 10000 billed chars and 64 KiB body; priced per character.
Only successful requests are billed. Verify prices and quotas in the official guide before relying on them.
Security notes
The API key is read from the environment and never logged or written to disk.
OCR/ASR read only the file path you pass; set
DEEPSEEK_TOOLS_ALLOWED_DIRSto confine reads.TTS writes only to
output_path(or the cache directory) and creates the parent directory.No shell execution, no network access beyond the configured
DEEPSEEK_TOOLS_BASE_URL.Responses are capped by
DEEPSEEK_TOOLS_MAX_OUTPUT_CHARSto protect the model's context.
Development
npm install
npm run build
npm test # unit tests (mocked fetch)
npm run smoke # live API calls; requires DEEPSEEK_API_KEYLive OCR / ASR checks:
SMOKE_OCR_FILE=/path/to/file.pdf \
SMOKE_ASR_FILE=/path/to/audio.wav \
DEEPSEEK_API_KEY=sk-... npm run smokeLicense
MIT
Available Tools
5 toolsdeepseek_asrDeepSeek ASRA
Transcribe a local WAV (16-bit PCM) or Ogg Opus audio file with DeepSeek's ASR API. Max 8 MiB / 120 seconds.
Inputs: file_path (required).
deepseek_fetchDeepSeek FetchA
Fetch a web page through DeepSeek's Fetch API and return its readable content as Markdown. Supports character-offset paging for long pages.
Inputs: url (required), offset, max_chars.
deepseek_ocrDeepSeek OCRA
Extract text from a local PDF, Word, PowerPoint or image file with DeepSeek's OCR API. Returns per-page text.
Inputs: file_path (required), purpose.
deepseek_searchDeepSeek Web SearchB
Search the web with DeepSeek's Search API. Returns titles, URLs, snippets/summaries, dates and source authority. Supports search operators and an original-question hint.
Inputs: query (required), count, localization, need_content, +1 more.
deepseek_ttsDeepSeek TTSA
Synthesize speech with DeepSeek's TTS API and write the audio to a local file. Returns the saved path, format and byte size.
Inputs: text (required), format, voice_id, output_path, +1 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 5 tools
Each tool maps to a clearly distinct capability: web search, page fetch, OCR, ASR, and TTS. There is no meaningful overlap between their purposes, so an agent can select the right tool from the task type alone.
All tools use a consistent deepseek_ prefix followed by a concise snake_case action or modality name. The pattern is predictable and easy to scan.
Five tools is well-scoped for a multimodal API wrapper covering search, retrieval, and audio/document processing. Each tool earns its place without redundancy.
The surface covers the core modalities promised by the server: search, fetch, OCR, ASR, and TTS. Minor gaps exist, such as no URL-based OCR/ASR or batch processing, but these are workaroundable and not central to the stated purpose.
Maintenance
Related MCP Connectors
OCR, transcription, file extraction, and image generation for AI agents via MCP.
Web search API and MCP server for developers building AI agents.
214 tools for AI agents over one MCP endpoint: image, video, audio, 3D, data, storage, mail, GPUs.
Scrape, crawl and search the web for AI agents via MCP.
Related MCP Servers
- AlicenseAqualityCmaintenanceExposes DeepSeek's Search and Fetch REST APIs as two read-only MCP tools, letting any MCP-capable agent run ranked web searches and retrieve individual pages as truncated Markdown. Works with whichever model the host runs, over stdio or Streamable HTTP.2MIT
- AlicenseAqualityCmaintenanceExposes DeepSeek's web search and page-fetch tools to any MCP client, letting any LLM run searches with quality signals like rerank score and source authority, plus advanced query syntax. It also retrieves single pages as Markdown, with built-in throttling, retries, and size-conscious summaries.2101 npm2MIT
- AlicenseAqualityCmaintenanceEnables web search and page fetching through the DeepSeek Search and Fetch APIs, exposing search and fetch tools to MCP clients. Supports paginated content retrieval with cursors so long pages can be read in chunks without repeating upstream requests.232 npmAGPL 3.0
- AlicenseAqualityCmaintenanceEnables MCP clients to run web searches that return ranked results with titles, URLs, snippets, summaries, and optional extracted page content, and to fetch HTTP or HTTPS pages as Markdown. It runs over stdio as a zero-dependency Python server that uses DeepSeek's official APIs and any supported MCP client.22MIT