Launching today

TranscriptLab
Fast YouTube captions & audio transcription for devs and AI
2 followers
Fast YouTube captions & audio transcription for devs and AI
2 followers
TranscriptLab turns any YouTube video or audio file into structured transcripts with timestamps. Search quotes, export to SRT/VTT/Markdown/JSON, and automate with our developer API or native MCP server for Claude, Cursor, and AI workflows. Free starter credits included.




Free Options
Launch Team

Render Sandboxes: a computer for every agentRun untrusted code safely. Idle sandboxes cost nothing
Promoted
I built TranscriptLab after burning credits on speech-to-text for videos that already had captions. It reads the caption track first and only falls back to transcription when there isn't one, so the common case is fast and nearly free.
The MCP server came later, when I realized that every time I was using Claude, Cursor, or ChatGPT to research a topic or learn from a tutorial, I was stuck copying YouTube links, opening external converters, and pasting messy subtitle text into the chat.
With Model Context Protocol (MCP), your AI assistant can directly fetch structured YouTube transcripts, search timestamps, and pull video context on demand—no manual copy-pasting needed.
Key features:
• Fast caption extraction first, optional STT audio fallback when needed
• Structured timestamp search & multi-format export (TXT, SRT, VTT, Markdown, JSON)
• Native MCP server & Claude plugin for instant AI assistant integration
• REST API + TypeScript SDK with async batch and channel jobs
How to try it:
Visit transcriptlab.dev to test the web extractor, or connect the MCP server directly to Claude / Cursor with free starter credits.
I'd love to hear your feedback:
1. What video platforms or formats should we support next?
2. How are you currently pulling video context into your AI workflows?
Ask me anything