spoken.md

API that fetches podcast transcripts as searchable Markdown with speaker names and timestamps

Audio & Voice TranscriptionSubscription

Visit website ↗

$15 / 100Starting price

6.5KMonthly visits

14Popularity

About spoken.md

spoken.md is an API that provides podcast transcripts as structured Markdown text. Users can search for episodes by title, guest, or a URL from Spotify, YouTube, or other podcast platforms. The API returns full transcripts with real speaker names and timestamps, avoiding generic labels like “Speaker 1.” Transcripts are provided in the original language of the recording, supporting multiple languages including Italian, German, Spanish, French, and Portuguese. The service operates on a credit system where one credit is consumed per episode fetched, regardless of length. Users can list all episodes of a show for free, then fetch transcripts individually. A scheduled polling endpoint helps keep archives current by listing new episodes since the last fetch. The API is designed for integration with knowledge bases, LLMs, or agent frameworks like Claude Code or Cursor.

Key features

Search episodes by title, guest, or URL
Fetch full transcripts as Markdown with speaker names
List all episodes of a show for free
Poll for new episodes since last fetch
Multi-language transcript support
Integration with agent frameworks via OpenAPI spec
MCP server for local clients
Real-time credit tracking and usage reporting

Use cases

Building searchable podcast archives for research or knowledge bases
Integrating podcast content into AI agents or LLMs
Automating transcription workflows for content creators

Pros

Returns transcripts as Markdown with resolved speaker names and timestamps
Supports multiple languages including Italian, German, Spanish, French, and Portuguese
Free search and metadata listing before any purchase
Credits never expire and errors are not charged
Designed for integration with LLMs and agent frameworks

Cons

No free tier for transcript fetching, only metadata
Limited to podcasts that publish transcripts
Requires API key and credit system for access
Availability varies by episode rather than by language

Frequently asked questions about spoken.md

What does spoken.md do?

spoken.md is a podcast transcript API that converts episodes into structured Markdown text with real speaker names and timestamps. It allows users to search for episodes by title, guest, or URL and fetch full transcripts in the original language of the recording.

Who is spoken.md designed for?

The tool is designed for developers, researchers, and AI agents who need searchable podcast archives for integration with knowledge bases, LLMs, or agent frameworks like Claude Code or Cursor.

How does the pricing model work?

Pricing is based on a credit system where one credit is consumed per episode fetched, regardless of length. Users can check the episode count of a show for free before purchasing credits, and credits never expire.

Does spoken.md support multiple languages?

Yes, transcripts are provided in the original language of the recording, supporting languages such as Italian, German, Spanish, French, and Portuguese. Availability varies by episode rather than language.

Can I integrate spoken.md with AI agents or LLMs?

Yes, spoken.md offers an OpenAPI spec, an MCP server, and an agent skill for frameworks like Claude Code, Cursor, and Windsurf, enabling seamless integration with AI agents and LLMs.

What happens if I fetch an episode I already pulled?

Fetching an episode you already pulled is never charged again, ensuring you only pay for new transcripts.