Open Models · first seen 13 October 2026, 23:27 UTC
Fish Audio releases Drama 3 open-weight TTS MoE model
A large open-weight TTS MoE expands accessible speech synthesis.
Drama 3 is a 35B-parameter open-weight mixture-of-experts text-to-speech model from Fish Audio; comparisons pit a preview against Qwen Audio 3.0 TTS.
Fish AudioOpenOpen weights
- Name
- Drama 3
- Params
- 35B