Video Content Extractor

One command to extract subtitles from any video — fully local, no API key.

---

Quick Start

pip install video-content-extractor
video-extract "https://www.douyin.com/video/7649006399359573937"

Commands

Command	Description
video-extract URL	Full pipeline: download → transcribe
video-extract file.mp4	Process local video
video-extract URL -l en	English transcription
video-extract URL --metadata-only	Info only, no download
video-extract URL -m medium	Higher quality model
video-extract URL -o ./out/	Custom output dir

Alias: vext works the same as video-extract.

Use in Other Agent Software

# Any agent that can run shell commands:
video-extract "https://www.youtube.com/watch?v=xxx" -l en -o ./output

Python API:

from video_extractor import Pipeline
pipe = Pipeline()
result = pipe.run_url("https://...", language="zh")
print(result["transcript"])

Requirements

• Python ≥ 3.10
• ~3 GB disk (Whisper model)
• ≥ 4 GB RAM

Supported Sites

Douyin, Bilibili, YouTube, and 1000+ more via yt-dlp.

License

MIT