MODEL REFERENCE / fishaudio
fishaudio/drama-3
Fish Audio's multi-speaker text-to-speech model (35B MoE, 3B active) served via vLLM-Omni, with voice cloning, speaker tags and inline event tags through the OpenAI /v1/audio/speech API.
1x H100: first audio 126 ms, RTF 0.28 at concurrency 1
View on HuggingFace
View on ModelScope02 / Configure
Configure serving
Choose engine, hardware and deployment options. The command updates as you go.
03 / Understand