Open-source local AI video analyzer powered by Ollama. Visual search, automatic highlights, scene/action/object detection, audio analysis, and subtitle generation. Free, offline alternative to Twelve Labs, Runway, and Descript.
An all-in-one, pure C++ inference engine for audio models, powered by ggml. Supports TTS, STT, VAD, voice conversion, music generation, and more, with highly optimized performance. No Python dependency.
Shared live audio from YouTube. Add tracks and playlists to the queue, stream to browsers and Sonos and more. - GitHub - 76696265636f646572/Airwave: Shared live audio from YouTube. Add tracks and playlists to the queue, stream to browsers and Sonos and more.
Open-source desktop video downloader powered by yt-dlp and electron. Download videos and audio from YouTube, Vimeo, Twitter, and thousands of sites. Works on macOS, Windows and Linux.