TranscriptFetch
Transcript data from any YouTube, TikTok or Instagram video, or a whole channel or playlist, inside your assistant. Analyze a video, a creator, or a whole topic.
About TranscriptFetch
TranscriptFetch is an API that converts video content from YouTube, TikTok, and Instagram into structured, timestamped text. It resolves a submitted video URL by checking a cache, reading published platform captions when available, or transcribing the audio directly when no caption track exists, returning results in a consistent JSON schema of per-segment text with start times, language, duration, and credit usage. It also supports resolving a channel or playlist into a paginated, newest-first list of videos, with polling via a since_video_id parameter that avoids charging credits when no new videos are found.
The service bills per credit for successful transcript fetches, with audio transcription billed per minute of audio processed, and charges nothing for failed requests. It provides a hosted MCP server and a prebuilt n8n node for use with clients such as Claude, ChatGPT, Cursor, and VS Code, alongside a REST API with cURL, Node, and Python examples. Access requires an API key, authenticated via a bearer token.
The product is aimed at developers and AI builders working with video as a data source, including use cases such as continuous monitoring of channels or playlists for new uploads, building normalized research datasets from interviews, reviews, demonstrations, and lectures, and giving AI agents source-grounded access to spoken content with exact quotes and timestamps. Stated figures on the page include over 1,700,000 transcripts fetched, a 99.82% API uptime over the last 30 days, and a 2.9-second average response time over the same period.
Key features
- Fetches timestamped transcripts from YouTube, TikTok, and Instagram videos
- Transcribes audio automatically when no caption track exists
- Lists videos from a channel or playlist for monitoring new uploads
- Monitors channels or playlists on a schedule to capture new uploads
- Reachable via hosted MCP server and prebuilt n8n node for AI agents
- Charges nothing for failed transcript requests
Frequently asked questions
What platforms does TranscriptFetch support?
It supports YouTube, TikTok, Instagram, and direct video or audio URLs.
How does billing work?
Requests are billed in credits, with 1 credit for a standard transcript and audio transcription billed per started minute; failed requests are not charged.
Can AI agents like Claude or ChatGPT use this?
Yes, TranscriptFetch offers a hosted MCP server and a prebuilt n8n node so Claude, ChatGPT, Cursor, or VS Code can fetch transcripts directly.
How does it handle videos without captions?
It transcribes the audio automatically when a video has no caption track, returning the same structured response shape.
Can I monitor a channel for new content?
Yes, you can follow a channel or playlist and TranscriptFetch will check for new uploads on a schedule and add transcripts to a searchable archive.
Webz.io News API
Webz.io News APIs provide global news data for search, monitoring, analytics, and AI applications, with powerful filters and natural-language search.
SerpApi - Search API
SerpApi's Search API makes it easy and fast to scrape Google and other search engines.
Show your product to thousands of developers
· 100k monthly pageviews
· 7k newsletter subscribers
You might also like
Roboflow Universe
AIPre-trained computer vision models.
SkyBiometry
AIFace Detection, Face Recognition and Face Grouping.
Spam Hunter
AIFree service to classify text as spam using ML.
Summarize Text with AI
AIThis API generates customizable summaries of text and web pages using AI.
Unplugg
AIForecasting API for timeseries data.
WolframAlpha
AIProvides specific answers to questions using data and algorithms.
