Universal speech-to-text inference library supporting 16+ model families with GPU acceleration
Unified C++ speech engine with 43 ASR backends and 48 TTS engines in one binary, zero Python dependencies
Lightweight C++ inference engine for diffusion models
