Video caption API
A caption API that finishes the video.
Import a clip, generate word-timed captions, apply explicit corrections, choose a style, request a preview and prepare a publish-ready export through MCP or a typed REST API.
weshippedtheupdate
How it works
Build a complete video caption workflow
Use goal-oriented operations instead of stitching transcription, timing and rendering together yourself.
- 01
Import and transcribe
Create a caption project from an upload target or supported source, then generate timed words.
- 02
Review and update
Read the current project version and apply explicit word or style changes.
- 03
Preview and export
Request a preview, receive a fixed quote and start a paid render only after approval.
Designed for agents and products
A reliable caption job, not just a transcript.
Typed operations
MCP tools and REST routes share product contracts for the same project workflow.
Durable tasks
Long-running transcription and render work is tracked as recoverable task state.
Approval before spend
A paid render is quoted first and requires the approved amount before it starts.
Works with your agent
Add ClipSubtitles to your agent.
Choose your client, install the server and sign in through your browser. No API key to manage.
Read the MCP docs →Install
Claude Code
claude mcp add --transport http clipsubtitles https://api.clipsubtitles.com/api/mcp- First use
- Sign in through your browser
- Endpoint
https://api.clipsubtitles.com/api/mcp
Captions in context
Made to look at home in the video.
Questions
What you need to know.
What can the video caption API do?
It can create a project, generate captions, read and update the current version, render a preview, prepare an export, track tasks and cancel work.
Can an AI agent use ClipSubtitles?
Yes. The same workflow is exposed through MCP for compatible agents and through the REST API for product integrations.
Does the API start paid renders automatically?
No. It returns an immutable quote first; the render starts only with explicit approval for that amount.
Give your agent a complete caption workflow.
Start with the API overview, then run the same flow your users will see.