LabscoMCP early access ↗
GtvarMCP SERVER

Yt Dlp

Download a video or pull just its audio through yt-dlp, and read a video's metadata without downloading anything.

by Gtvar ↗Page Fetching & Content ExtractionCommunity

LABSCO SUMMARY

Metadata first, so the download is the one you meant.

get_video_info is the tool that earns its place: reading duration, uploader and available formats before committing to a download means a two-hour stream or a format your player cannot open is caught before the bytes move. Everything runs through your own yt-dlp install, so what it can reach is whatever that version supports.

What it is

A wrapper over yt-dlp, the command-line downloader, exposing it as two tools. One reads metadata only; the other downloads, with quality, container format and output path as arguments, and an audio-only switch that turns the same call into an extraction.

What you get

  • A video's metadata — title, duration, uploader and available formats — read without downloading it
  • A video downloaded to a location you choose, with quality and container preferences
  • Audio extracted instead, in mp3, m4a, ogg or opus
  • Output paths set per call through a filename template

Requirements

Yt-dlp installed on the machine — brew install yt-dlp on macOS, pip install yt-dlp on Linux — and npx on your PATH. Three optional environment variables: YTDLP_PATH points at yt-dlp if it is not on your PATH, YTDL_OUTPUT_DIR sets where files land (/tmp if unset), and FFMPEG_PATH points at ffmpeg.

Setup effort

One command — npx @gtvar/yt-dlp-mcp