Skip to content
MCP server

SpeechWeave

Transcribe local audio files and public URLs from your editor through SpeechWeave

Runs locallyStatus and history
Install
Pricing
Open source
Platforms
CLI, API
Runs
Locally
Version
1.10.0
Listing updated
2 days ago

About SpeechWeave

SpeechWeave is an open-source MCP server for ai models and llm tooling. It runs locally on your machine and is started by your AI client.

Eight tools cover both ways of getting a transcript. The sync pair, transcribe_file and transcribe_url, return text in the same call. The async pair, start_job_file and start_job_url, hand back a job id straight away so a long recording can be polled later with get_job_status or stopped with cancel_job. Audio is never passed as raw bytes: local clients give an absolute path and others give a public HTTPS URL.

Options include a core or max model, standard or deferred service mode (deferred costs less and has lower priority), a forced language code, translation to English, and a vocabulary prompt for names and acronyms. Transcripts can come back as text, SRT, VTT or verbose JSON, or be written to a .txt, .md, .json, .srt or .vtt file so long output stays out of the conversation.

Billing problems such as a spend cap or an empty wallet come back as tool errors with a hint telling the assistant not to retry. It needs Node.js 18+ and a SpeechWeave API key.

Install

Run in a terminal

claude mcp add mcp --env SPEECHWEAVE_API_KEY='<SPEECHWEAVE_API_KEY>' -- npx -y @speechweave/mcp

Add to claude_desktop_config.json

{
  "mcpServers": {
    "mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@speechweave/mcp"
      ],
      "env": {
        "SPEECHWEAVE_API_KEY": "<SPEECHWEAVE_API_KEY>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@speechweave/mcp"
      ],
      "env": {
        "SPEECHWEAVE_API_KEY": "<SPEECHWEAVE_API_KEY>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "mcp": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@speechweave/mcp"
      ],
      "env": {
        "SPEECHWEAVE_API_KEY": "<SPEECHWEAVE_API_KEY>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@speechweave/mcp"
      ],
      "env": {
        "SPEECHWEAVE_API_KEY": "<SPEECHWEAVE_API_KEY>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@speechweave/mcp"
      ],
      "env": {
        "SPEECHWEAVE_API_KEY": "<SPEECHWEAVE_API_KEY>"
      }
    }
  }
}

Configuration

  • SPEECHWEAVE_API_KEYSpeechWeave API key (sk_live_...)Required

Tools this server provides (8)

Taken from the project documentation; the running server may differ.

transcribe_file
Transcribes a local file and returns the text
transcribe_url
Transcribes a public HTTPS URL and returns the text
start_job_file
Starts an async job for a local file
start_job_url
Starts an async job for a URL
get_job_status
Polls status or transcript for a job id
cancel_job
Cancels a queued or processing job
get_limits
Shows upload size ceilings for the API key
fetch_doc
Fetches public docs by slug

What it does

Sync and async delivery
Get the transcript in one call or start a job and poll for it.
Output formats
Return plain text, SRT, VTT or verbose JSON with segment timings.
Save to file
Write the transcript to a file so long text stays out of the chat.
Translate or hint
Translate to English or pass a vocabulary prompt for proper nouns.
Billing hints
Spend-cap and balance errors include a hint so the assistant does not retry.
Limits lookup
get_limits reports upload size ceilings for your API key.

Good for

  • Transcribe a meeting recording from a path in your project
  • Turn a public podcast URL into SRT subtitles
  • Run a long file as a background job and check on it later
  • Translate foreign-language audio to English text

Pricing

Open source

Server is MIT-licensed; transcription needs a SpeechWeave API key with usage billing (prices not published here)

Prices change. Check speechweave.com before you buy.

SpeechWeave FAQ

How do I install it?

Set SPEECHWEAVE_API_KEY and run npx -y @speechweave/mcp. It needs Node.js 18 or newer.

Can I send raw audio over MCP?

No. Pass an absolute local file path or a public HTTPS URL.

What is deferred service mode?

A lower-cost, lower-priority mode that works with either sync or async tools.

What licence is it under?

The README badge shows the MIT licence.

Is SpeechWeave free?

Yes. SpeechWeave is open source, so you can use and self-host it for free.

How do I add SpeechWeave to Claude Code?

Run this command in your terminal: claude mcp add mcp --env SPEECHWEAVE_API_KEY='<SPEECHWEAVE_API_KEY>' -- npx -y @speechweave/mcp

What are the best alternatives to SpeechWeave?

The closest alternatives to SpeechWeave are Claude Code Ultimate Guide, ASTra, CARTESIAN ECU, Invoices and SigPen.

Alternatives to SpeechWeave

  1. MCP server for searching the Claude Code Ultimate Guide from any MCP client

    Runs locally
    • Open source
    • 6.5K GitHub stars
    • AI models and LLM tooling
  2. ASTra

    MCP server

    Let agents read, edit, validate and undo Word .docx files with exact operations

    Runs locally
    • 5 tools
    • Freemium
    • Files and documents
  3. CARTESIAN ECU

    MCP server

    Check and repair ECU and TCU flash-file checksums from Claude (Ultimate plan)

    Not checked yet
    • 5 tools
    • Remote
    • Paid
    • Files and documents
  4. Invoices

    MCP server

    MCP server that reads supplier invoices and returns data plus DATEV, CSV or Excel files

    • 4 tools
    • Remote
    • Paid
    • Files and documents
  5. SigPen

    MCP server

    MCP server that drafts e-signature documents while you approve every send

    Not checked yet
    • 13 tools
    • Remote
    • Freemium
    • Files and documents
  6. Transcribe audio and video links from Claude or ChatGPT and read the results

    Not checked yet
    • 5 tools
    • Remote
    • Freemium
    • Design and media