getraw

Fast media downloader CLI built natively in Bun/TypeScript. A yt-dlp replacement with native JS execution.

Why getraw?

• Native JS execution — YouTube's player code runs natively in Bun. No external runtime needed (yt-dlp requires Deno/Node).
• 50ms cold startup — Bun-powered, not Python.
• 30+ sites — YouTube, Twitter, TikTok, Instagram, Reddit, Twitch, and more.
• Zero API keys — All extractors use public endpoints, guest tokens, and page scraping.
• Agent-ready — Install as an AI agent skill: npx skills add onkits/getraw

Installation

bun install -g getraw

From source

git clone https://github.com/onkits/getraw
cd getraw
bun install

As an AI agent skill

npx skills add onkits/getraw

Works with Claude Code, Cursor, Copilot, Codex, Windsurf, and 50+ other agents.

Quick Start

# Download a video
getraw https://www.youtube.com/watch?v=dQw4w9WgXcQ

# Extract audio as MP3
getraw -x --audio-format mp3 https://soundcloud.com/artist/track

# List available formats
getraw -F https://vimeo.com/123456789

# Download specific quality with subtitles
getraw -f "bestvideo[height<=1080]+bestaudio" --write-subs https://www.youtube.com/watch?v=dQw4w9WgXcQ

# Get metadata as JSON (no download)
getraw -j https://www.reddit.com/r/videos/comments/abc123/some_post/

CLI Reference

Usage: getraw [OPTIONS] URL [URL...]

| Flag | Short | Default | Description | |------|-------|---------|-------------| | --format | -f | bv*+ba/b | Format selection string | | --output | -o | %(title)s [%(id)s].%(ext)s | Output filename template | | --extract-audio | -x | | Extract audio only | | --audio-format | | mp3 | Audio format (mp3, aac, flac, wav, opus) | | --write-subs | | | Write subtitles to file | | --sub-langs | | en | Subtitle languages | | --list-formats | -F | | List available formats | | --dump-json | -j | | Dump info JSON to stdout | | --quiet | -q | | Suppress output | | --verbose | -v | | Verbose output | | --retries | -R | 3 | Number of retries | | --rate-limit | -r | | Rate limit in bytes/sec | | --proxy | | | Proxy URL | | --cookies | | | Cookie file path (Netscape format) | | --embed-thumbnail | | | Embed thumbnail in output | | --embed-subs | | | Embed subtitles in output | | --version | -V | | Print version | | --help | -h | | Show help |

Supported Sites (30+)

| Site | URL Patterns | |------|-------------| | YouTube | youtube.com, youtu.be, shorts, live, playlists, channels | | Twitter/X | twitter.com/*/status/*, x.com/*/status/*, Spaces | | TikTok | tiktok.com/@*/video/*, vm.tiktok.com, user profiles | | Instagram | instagram.com/p/*, /reel/*, /reels/ | | Reddit | reddit.com/r/*/comments/*, v.redd.it, galleries | | Twitch | VODs, clips, live streams | | Vimeo | vimeo.com/*, player embeds | | SoundCloud | Tracks, playlists, albums | | Bilibili | Videos, bangumi/anime | | Dailymotion | Videos | | Bandcamp | Tracks, albums | | Kick | VODs, clips, live | | Rumble | Videos | | TED | Talks (with multi-language subtitles) | | Niconico | Videos | | Streamable | Videos | | Imgur | Videos, GIFs, albums | | Coub | Videos (video + audio merge) | | Odysee/LBRY | Videos | | PeerTube | Any instance | | Spotify | Podcast episodes (30s preview) | | Archive.org | Any public media | | Google Drive | Public files | | Dropbox | Public share links | | + more | Generic fallback for direct media URLs |

See docs/supported-sites.md for full details.

For AI Agents

getraw is designed to be used by AI agents. Key commands for automation:

# Get structured metadata
getraw --dump-json "URL" | jq '.title, .duration, .formats[0].url'

# Download transcript for summarization
getraw --write-subs --sub-langs en --skip-download "URL"

# Extract audio for transcription pipelines
getraw -x --audio-format wav -o "audio.wav" "URL"

# Batch download
getraw URL1 URL2 URL3

Install as an agent skill for any compatible AI coding agent:

npx skills add onkits/getraw

Building from Source

git clone https://github.com/onkits/getraw
cd getraw
bun install
bun test         # 386 tests
bun run build    # standalone binary

Writing a Custom Extractor

See docs/plugin-guide.md for the BaseExtractor interface and examples.

License

MIT