1B parameter high-fidelity audio generation model.
High-performance automatic speech recognition model supporting 60+ languages with OpenAI Whisper-compatible API
