Description

TranscriptFetch is a transcript API for developers. Send a YouTube, TikTok, Instagram, Spotify, Apple Podcasts or RSS episode link and get back the full text plus timestamped segments in a single call.

Most transcript tools break on the two hard cases. TranscriptFetch handles both. When a video has no caption track, which is the majority of TikToks and Reels, it transcribes the audio instead. When a podcast link points at a directory rather than a file, it resolves the episode to the publisher's real audio and returns the show and episode names with the text.

Built for production use: a REST API, official Node and Python SDKs, and a Model Context Protocol server so AI agents in Claude, Cursor and other MCP clients can fetch transcripts directly. Responses are cached, so a repeat request for a popular video returns in milliseconds.

Pricing is per successful response rather than per audio minute, which makes long content cheap. Failed requests cost nothing. Every account gets 100 free credits every month, with paid plans from $5.