Balabolka

Hugging Face distilled version of OpenAI Whisper. 6x faster, 49% smaller, with similar accuracy for speech recognition.

Open-source text-to-speech library. Train and use TTS models with support for multiple languages and voices.

Suno AI open-source text-to-audio model. Generates speech, music, sound effects, and nonverbal sounds from text.

Style-based TTS model that generates natural speech with expressive style control. Zero-shot voice cloning,

Fast speech-to-text model optimized for edge devices. Real-time transcription with low latency and small footprint.

Zero-shot voice cloning and text-to-speech. Clone any voice from just a few seconds of audio.