Promote Sign in Sign up

Audio model

Higgs Audio v3 TTS

Higgs Audio v3 TTS is Boson AI’s text-to-speech model for expressive conversational voice agents across 100+ languages with zero-shot voice cloning and inline speech controls.

Boson AI Audio Open weights other
Visit model website ↗ Hugging Face
Type
Audio
Weights
Open
Downloads
12
Pipeline
text to speech
License
other
Library
transformers

Model hub

Umn244/higgs-audio-v3-tts-4b

Repo
Umn244/higgs-audio-v3-tts-4b
Access
Public
Technical tags and languages
af ar as az ba be bg bn bs ca cs transformers safetensors higgs_multimodal_qwen3 text-generation text-to-speech speech-generation voice-agent expressive-speech controllable-tts multilingual-tts ast ceb
Open on Hugging Face

More from Boson AI

3

About Boson AI

1

Boson AI builds voice agents using proprietary multimodal foundation models. Its Higgs Audio family enables expressive, multilingual text-to-speech and speech-to-text across 100+ languages, with zero-shot voice cloning and inline prosody control. Feynman Flow is its agentic platform for deploying voice intelligence in enterprise workflows.