Gaziyav models
The on-device AI models of Gaziyav, the video editor: speech to text (Whisper), object removal (LaMa, SAM 2), voice clean-up, voice and music separation, voice changer and retouch. The app downloads each file the first time a feature is used and checks it against MANIFEST.txt (size and SHA-256); the authors' original files stay the fallback.
Each model keeps its authors' licence: the LICENSE* file in its folder.
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐ Ask for provider support