Video
AI-Video-Summarizer
YouTube summaries and timestamps via Gemini or GPT.
★ 98 starsMIT
View guide →Self-hosted AI video analysis: summaries, chapters, transcripts.
koma is an open-source audio project by yuxino-labs (yuxino-labs/koma on GitHub). It has 236 GitHub stars and 2 forks, is released under the Custom license, and was last updated on GitHub on 22 September 2026.
Self-host it per the official site.
Feed in the footage to analyze.
Summaries, chapters, keyframes, transcript.
Take SRT or JSON into your workflow.
AI-written summaries of footage.
Chapter markers from content.
Visual highlights pulled automatically.
Full speech-to-text output.
Structured data for your pipeline.
Runs on your own hardware.
Download the source and enter the folder.
git clone https://github.com/yuxino-labs/koma.git
cd komaSet up the Python environment.
pip install -r requirements.txtLaunch it per the README.
python app.pySelf-hosted AI video analysis with summaries, chapters, keyframes, and transcripts, at yuxino-labs/koma.
Summary, chapters, keyframes, transcript, plus SRT/JSON export.
The code is on GitHub under a custom license; check the terms.
Yes. It is designed to be self-hosted.
Computer vision, speech-to-text, and ffmpeg.
YouTube summaries and timestamps via Gemini or GPT.
MCP server that turns any video — YouTube.
Ai-video-captions, in TypeScript — built for free, open-source ai video caption generat...
Self-hosted backend that gives ESP32 'Xiaozhi' devices streaming voice chat, memory and tool control.
Nebulynk — a self-hosted home for your team's conversations, calls, publicly available.
Self-hosted Whisper speech-to-text as a Docker image.
The most-starred other Audio projects in the LibreAI catalogue — the realistic shortlist if koma is not the right fit.
利用 AI 大模型和自动化工作流，根据主题或关键词一键生成高清短视频。Generate HD short videos from a topic or keyword with an automated AI workflow.
Robust Speech Recognition via Large-Scale Weak Supervision
1 min voice data can also be used to train a good TTS model! (few shot voice cloning)
Clone a voice from 5 seconds of audio and generate arbitrary speech in real time
Port of OpenAI's Whisper model in C/C++
Free and open — grab the source, star the repo, and start creating.