How does authentication work?
Authentication is via the `x-api-key` header. A demo key (`pt_demo`) is available for testing (search works fully; transcripts limited to a demo episode).
Spoken MCP Server
@roberttomko
Basic information
About Spoken MCP Server
Podcast transcript API — turn any published podcast into clean Markdown with real speaker names. Built for AI agents.
Spoken MCP Server config
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"spoken": {
"command": "npx",
"args": [
"-y",
"spoken-mcp"
],
"env": {
"SPOKEN_API_KEY": "pt_demo"
}
}
}
}Spoken MCP Server tools
4Search published podcast episodes by text query, or paste an episode URL (Spotify, YouTube, etc.). Returns matching episodes with their id, title, podcast, podcast_id, and date. Use the id with get_transcript, or the podcast_id with list_episodes to get the show's whole back-catalog. Does not consume credits.
Fetch a podcast episode's transcript as clean Markdown with real speaker names and timestamps. Pass an episode id from search_podcasts. Costs 1 credit on the first fetch of an episode; repeat fetches are free and errors are never charged.
List a podcast's entire back-catalog (every episode, newest first). Pass a podcast_id from a search_podcasts result. Returns each episode's id, title, and date — fetch any with get_transcript. Use this to transcribe a whole show. Does not consume credits itself; transcribing the returned episodes costs 1 credit each (repeat fetches are free), so make sure the key has enough credits before looping.
Check the current Spoken credit balance, account email, and recent usage for the configured API key. Does not consume credits.
Overview
What is Spoken?
Spoken is a transcript API that turns any published podcast into clean Markdown with real speaker names, not “Speaker 1.” It retrieves named, timestamped text via a single API call, designed for LLMs, RAG pipelines, summarizers, and search. It’s a retrieval service for already-published shows, typically 5–10× cheaper than running audio through a transcription service.
How to use Spoken?
Install and configure the MCP server via npx -y spoken-mcp or from source, setting the SPOKEN_API_KEY environment variable. Add the server to your MCP client config (e.g., Claude Desktop’s claude_desktop_config.json) to expose tools like search_podcasts, list_episodes, get_transcript, and get_balance. Alternatively, use the HTTP API directly via curl with an x-api-key header.
Key features of Spoken
- Real speaker names resolved automatically
- Clean Markdown with timestamps tuned for LLMs
- Search by text query or paste Spotify/YouTube URL
- Pay-per-use credits with no subscription
- Repeat fetches are free after first charge
- Agent-native with MCP server and OpenAPI spec
Use cases of Spoken
- Fetch a podcast transcript and summarize it with an LLM
- Chunk transcripts for a vector store or RAG pipeline
- Search for episodes by topic or podcast name
- Integrate as a tool in MCP-compatible agents (Claude, Cursor, Cline)
Frequently asked questions
Is there a free tier or demo?
Yes, you can use `pt_demo` as the API key to try search and get one transcript from the demo episode. Full access requires a paid key from spoken.md.
What does it cost?
Spoken uses pay-per-use credits: 100 for $15, 500 for $50, 2,000 for $160. Failed calls are never charged, and repeated fetches of the same transcript are free after the first request.
Can I use it with any podcast?
Spoken works on already-published podcasts. You can find episodes by text search or by pasting a Spotify or YouTube URL. It does not process uploaded audio.
What happens if I run out of credits?
The API returns credit balance info in the `X-Credits-Remaining` header. You can top up or buy a new key via Stripe endpoints (`/buy`, `/top-up`).
More Media & Design MCP servers
LinkedIn Ads MCP
AdPlugLinkedIn Ads MCP server by AdPlug, connecting LinkedIn Ads to Claude, ChatGPT and other MCP-compatible AI assistants. Analyse performance, audiences, creatives and conversions. Generate reports and manage campaigns, budg
Aura AI
auraai.appAccess 20+ AI video and image models from Google, OpenAI, xAI, ByteDance, and more. Generate and edit videos with Veo 3.1, Sora 2, Seedance, Kling O3, and images with FLUX, Nano Banana, Seedram 4, GPT Image — all in one
Veltake Studio
Veltake AIDrive Veltake Studio from any AI agent: create film and ad projects, have the Studio agent build scenes on the canvas, price and start renders, export edited cuts, and make single images, video, speech and music in one c
GOJI
GOJI AgencyA remote MCP server from GOJI, a full-service digital agency in Melbourne, Australia. Nine tools over published material: a 104-term plain-English glossary, a 14-chapter guide to AI visibility, seven long-form guides inc
SnapDub
SnapDubTurn long videos into captioned vertical clips for TikTok, Reels and Shorts from Claude, ChatGPT, Cursor or any MCP client. Paste a YouTube, Twitch, Kick, Rumble or Zoom link and get clips with titles, virality scores, d
Comments