1 min voice data can also be used to train a good TTS model! (few shot voice cloning)

A fast multimodal LLM for real-time voice

Orpheus TTS Server with streaming support (TTFB ~160ms)

Chatterbox with OpenAI-compatible endpoints, streaming support, multiple voices, and live LLM chat integration. Docker-ready with GPU acceleration.

Medibeng-Orpheus-3b-0.1-ft- A TTS model for bilingual Bengali-English code-switching in healthcare, fine-tuned for seamless patient-doctor interactions.

AI-powered real-time voice interaction framework for building conversational audio agents. Modular Python toolkit for streaming Speech-to-Text (OpenAI), Text-to-Speech (Orpheus 3B), and real-time a…