Skip to content
MCP server

Spoken

Podcast transcripts as Markdown with real speaker names, searchable by agents and scripts

Pricing
Freemium
Platforms
API, CLI, Web
Runs
Locally or Remotely
Version
0.4.3
Listing updated
1 day ago

About Spoken

Spoken is a freemium MCP server for news and media. It can run locally on your machine or as a hosted remote server.

Spoken is a retrieval service, not speech-to-text: it works on already-published podcasts, so no audio is uploaded. Search by title, guest or a pasted Spotify or YouTube link; list every episode of a show with a transcript in one free call; then fetch each transcript as timestamped Markdown where speakers carry resolved names rather than Speaker 1 and Speaker 2. A /new endpoint lists episodes you have not fetched yet, with an Atom feed option, so a scheduled job can keep a notes archive current.

Pricing is per transcript through credits that never expire: one credit per episode, charged on first fetch only, with repeat fetches and failed calls free. Packs are listed at $0.15, $0.10 and $0.08 per transcript, and a full refund is offered in the first 14 days if 20 or fewer credits were used. The demo key pt_demo makes search work fully but limits transcripts to a demo episode.

The MCP server runs locally over stdio through the spoken-mcp npm package or as a hosted endpoint, and there is also an Agent Skill, an OpenAPI spec and a Python package.

Install

Run in a terminal

claude mcp add --transport http spoken https://spoken.md/mcp

Add to Settings → Connectors → Add custom connector

https://spoken.md/mcp
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "spoken": {
      "url": "https://spoken.md/mcp"
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "spoken": {
      "type": "http",
      "url": "https://spoken.md/mcp"
    }
  }
}

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

{
  "mcpServers": {
    "spoken": {
      "serverUrl": "https://spoken.md/mcp"
    }
  }
}

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

{
  "mcpServers": {
    "spoken": {
      "type": "http",
      "url": "https://spoken.md/mcp"
    }
  }
}

Configuration

  • SPOKEN_API_KEYAPI key from spoken.md (starts with pt_). Defaults to pt_demo (search works fully; transcripts limited to the demo episode).

What it does

Real speaker names
Speakers are identified from context instead of numbered labels.
Clean Markdown
Timestamped text shaped for LLM context windows and RAG chunking.
Back-catalog listing
Listing a show's episodes is free.
New-episode polling
/new returns unfetched episodes, optionally as an Atom feed.
Pay per transcript
Credits never expire and repeat fetches are free.
Demo key
pt_demo lets search work fully before you buy.

Good for

  • Archive a show's full back catalog into a knowledge base
  • Summarize a podcast episode with an assistant
  • Find the episode where a guest discussed a topic
  • Chunk transcripts for a RAG pipeline

Pricing

Freemium

Pay-per-transcript credits ($0.08 to $0.15 each); free demo key

Prices change. Check spoken.md before you buy.

Spoken FAQ

How is it priced?

Credit packs cost $0.15, $0.10 or $0.08 per transcript; one credit fetches one episode.

Is a repeat fetch charged?

No. Fetching an episode you already pulled is free, and failed calls are not charged.

Can I try it free?

Yes, with the demo key; search works fully but transcripts are limited to one demo episode.

Does it transcribe my own audio?

No. It retrieves transcripts for already-published podcasts.

Is Spoken free?

Spoken has a free plan, with paid plans for more usage or features.

How do I add Spoken to Claude Code?

Run this command in your terminal: claude mcp add --transport http spoken https://spoken.md/mcp

What are the best alternatives to Spoken?

The closest alternatives to Spoken are Codebase Memory, XERJ, Claude Code Ultimate Guide, Nautilus Compass Memory and note1.

Alternatives to Spoken

  1. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  2. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  3. 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
  4. Local-first long-term memory layer for AI agents, served over MCP

    Not checked yet
    • 5 tools
    • Remote
    • Open source
    • 1.3K GitHub stars
    • Knowledge and memory
  5. note1

    MCP server

    MCP server giving AI assistants search and summaries across your note1 meetings

    • 4 tools
    • Remote
    • Free trial
    • 1 GitHub star
    • Productivity