Extract clean text from YouTube, TikTok, Facebook, Instagram, and 7+ more. Auto-detects the fastest method: API → subtitles → Whisper fallback.
yt-transcript extracts text transcripts from video on 11+ platforms via an auto-fallback pipeline: YouTube transcript API (~2s) → yt-dlp subtitle extraction (~10s) → faster-whisper GPU (~5–15 min). Two scripts: universal_transcript.py (all platforms + Whisper) and yt_transcript.py (YouTube, playlist, channel).
- 11+ platforms: YouTube, TikTok, Facebook, Instagram, Twitter/X, Bilibili, Twitch, Vimeo, Dailymotion, Reddit, Douyin
- Auto-fallback: API → subtitles → Whisper GPU
- Output: txt, srt, json in
./transcripts/ - YouTube: playlist + channel batch support
pip install youtube-transcript-api faster-whisper
# also: yt-dlp + ffmpeg on PATH
python scripts/universal_transcript.py "URL" --lang vi
python scripts/yt_transcript.py "PLAYLIST_URL" --lang vivideo transcript, lấy nội dung, transcript, lấy text từ video, subtitle, phụ đề, extract text from video
- tg-voice-whisper — voice message transcription
Part of the NachaFromMars OpenClaw skill ecosystem.