ringger/transcribe-critic

Multi-source transcript merging inspired by textual criticism — LLM adjudicates multiple Whisper, YouTube captions & external transcripts for higher quality. Includes speaker diarization and summarization.

41
/ 100
Emerging

This tool helps you create highly accurate written transcripts from video or audio recordings, especially for lectures, interviews, or presentations. It takes a video URL (like YouTube) or audio podcast link, along with any existing captions or transcripts, and produces a single, meticulously reviewed transcript. Content creators, researchers, journalists, or anyone needing precise text from spoken content would find this invaluable.

Available on PyPI.

Use this if you need the most accurate possible transcript from a recording, especially when multiple sources (like YouTube captions and automatic speech recognition) are available.

Not ideal if you only need a quick, basic transcript from a single source and speed is your top priority over absolute accuracy.

transcription content-creation research journalism media-analysis
Maintenance 10 / 25
Adoption 5 / 25
Maturity 20 / 25
Community 6 / 25

How are scores calculated?

Stars

14

Forks

1

Language

Python

License

MIT

Last pushed

Mar 09, 2026

Commits (30d)

0

Dependencies

7

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/ringger/transcribe-critic"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.