VideoLingo
Open-source pipeline for Netflix-grade subtitles, translation, and dubbing.
About
VideoLingo is an open-source project that automates video localization end to end: word-level transcription with WhisperX, NLP and GPT-based subtitle segmentation, multi-step translation, and TTS dubbing. Its signature output is Netflix-standard single-line subtitles, and it runs as a one-click Streamlit app with detailed logging and progress resumption.
Licensed under Apache-2.0 with a large community (reviewer-reported 18,000+ GitHub stars), it supports local installation, Docker, batch mode, and multiple dubbing backends including IndexTTS2 voice cloning. It can run fully locally with no API cost.
What it can do
- Word-level subtitle recognition
- GPT-based subtitle segmentation and translation
- Netflix-standard single-line subtitles
- TTS dubbing with voice cloning
- Batch processing mode
Setup
Use the one-click package, Docker image, or source install from GitHub; configure LLM and TTS backends per the docs.
Prerequisites
Python environment or Docker; optional LLM API key or local models for translation.
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 116; Trending score 76/100 from researcher assessment.
★ 18,000 GitHub stars
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.