2. 15+ Inline Emotion Tags for Expressive Speech

3. Streaming Audio Generation

4. vLLM integration

5. Runs on single GPU

https://huggingface.co/maya-research/maya1