AIMarkTechPostabout 4 hours ago
Fish Audio Releases Drama 3 TTS Model
1 min read
Fish Audio released Drama 3, a 35B Open-Weight MoE text-to-speech model with about 3B active parameters per token. It runs on one NVIDIA GPU with at least 48 GB in FP8 and is available on Hugging Face and the Fish Audio API for scripted dialogue and dubbing.
Level
Hype check
Tap to vote and see what everyone thinks.
