Repository navigation
Dub Studio 4.5.1
Every engine runs on any x86-64 processor, with or without AVX2. The audio.cpp server (Higgs, Supertonic, CrisperWhisper, Parakeet, diarization), recognition (transcribe.cpp), Qwen3-TTS and the voice separator (BSRoformer) now ship a processor build for every instruction level and load the one this processor runs best, instead of the AVX2 build alone; 4.5.0 refused these engines on a processor without AVX2. The new builds are downloaded once after the update.
The Linux packages are experimental. macOS builds follow by tag.
See the CHANGELOG for the full list.