Subtitles and metadata only

MCP tools return subtitle text and metadata only — extracted via yt-dlp from public URLs. Video and audio files are not delivered to the client and are not intended for end-user download through this product.

Public content only

We work with publicly accessible URLs on supported platforms. No bypass of DRM, paywalls, or private accounts.

No training on your data

Transcripts from your requests are not used to train Video Intelligence MCP models. Hosted retention policy will be in the final Privacy Policy.

Responsible use

You are responsible for complying with source platform ToS and applicable law when analyzing content.

Source & stack

transcriptor-mcp on GitHub
Open-source MCP server — Docker image, tool schemas, and self-hosted setup.
yt-dlp
Upstream extractor for public subtitles and video metadata (used under the hood).

Security and DPA questions: security@video-intelligence.dev (placeholder).

Join early access