TranscriptFetch MCP
officialVideo and audio transcripts for AI agents: YouTube, TikTok, Instagram, podcasts, plus video search.
What can you do with TranscriptFetch MCP?
- Fetch video transcripts — Ask your assistant to pull captions from any YouTube, TikTok, or Instagram link via
https://transcriptfetch.com/mcp. - Get timestamped quotes — Request a transcript with start times per segment so the AI can cite exact moments from a clip.
- Handle missing captions automatically — When a video has no captions, the server triggers AI Fallback Transcription; the assistant can retry to collect the text.
- Use one format across platforms — Get the same response fields for YouTube, TikTok, and Instagram links, so no branching logic is needed.
- Track usage per client — Give each client or agent its own key with separate logs, and revoke access individually without affecting others.
Documentation
Your assistant stops at the play button
Ask an assistant about an article and it answers. Send it a video and it guesses from the title, the description, whatever the page happened to expose. The words spoken in the clip are the part it never sees.
Without
TitleDescriptionWhatever the page exposed
With TranscriptFetch
00:12Most of our corpus was video,00:19which meant most of it was not00:24searchable at all…
Everything the server handles for you
Fallback on every fetch
Captions come back immediately. When a video has none, AI Fallback Transcription starts as a job and the assistant calls the tool again to collect the text.
Timestamps on every segment
Each line carries its start time, so an assistant can cite the moment instead of summarising the whole thing.
One shape, every platform
A TikTok link and a YouTube link return the same fields. Nothing downstream has to branch on where the link came from.
Failures cost nothing
A credit is spent when text comes back. Failed fetches, private videos and dead links are never billed.
A key per client
Separate keys per client or per agent, each with its own log, so you can see what asked for what and revoke one alone.
No infrastructure to run
No queue to babysit, no proxy pool, no scraper to repair the week a platform changes its markup.
Set up your client
Server URL · streamable HTTP https://transcriptfetch.com/mcp
By client
Claude MCP server setup
All three Claude paths on one page: the claude.ai one-click connector (OAuth, no key), Claude Code's one-liner, and the Claude Desktop mcp-remote bridge.
Cursor MCP server setup
The ~/.cursor/mcp.json entry, and the same block for VS Code, so the tools appear in Agent mode after a restart.
Codex MCP server setup
Codex configures MCP over stdio, so this is the mcp-remote bridge and the ~/.codex/config.toml it belongs in.
By platform
YouTube MCP server
Ask about any YouTube link in the chat. Captions where the video publishes them, AI Fallback Transcription where it does not.
TikTok MCP server
Short-form clips read as text, with timestamps, so an assistant can quote what was said rather than the caption.
Instagram MCP server
Reels and video posts transcribed on demand, returned in the same shape as every other source.
Frequently asked questions
What is MCP, in one sentence?
A standard way for an AI client to call outside tools. You add our server once and your assistant gains the ability to fetch transcripts, the same way it might gain a calendar or a file system.
Do I need to write any code?
No. Adding the server is pasting a URL and a key into your client's settings. Code is only involved if you are building your own agent, in which case the same server works through the API.
What happens when a video has no captions?
AI Fallback Transcription kicks in automatically - short videos come back in the same call, typically in ~30 seconds. Long media becomes a background job the assistant collects on its next call, or it retries with ai_fallback: true when the response suggests it.
How is this billed?
Caption fetches cost one credit; AI Fallback Transcription is billed by audio duration, only on delivery. Both draw from the same allowance as the API. Failures, private videos and blocked or unavailable links cost nothing.
Can I control what my assistant can reach?
Yes. Each key is logged separately, so you can give one client its own key, watch what it calls, and revoke it without touching the others.
Does it work with clients that are not on your list?
If it speaks MCP, it works. Nothing in the server is specific to a particular client, so anything that ships support later will connect with the same URL.