Zen3 TTS

Canonical Zen3 text-to-speech model. 12 Hz, 1.7B parameter base voice generator producing natural prosody across English and Chinese for voice agents and media production in the Zen3 audio family.

Hosted via Hanzo gateway

Served at api.hanzo.ai as zen3-tts.

Weights

Mirror staging in progress. Use the canonical base via the base_model link in frontmatter until weights land here.

Acknowledgements

Built on Qwen/Qwen3-TTS-12Hz-1.7B-Base (Apache-2.0). Mirrored here for the Zen3 audio family.

Downloads last month: 22

Safetensors

Model size

2B params

Tensor type

BF16

Model tree for zenlm/zen-3-tts

Base model

Qwen/Qwen3-TTS-12Hz-1.7B-Base

Finetuned

(25)

this model

Collection including zenlm/zen-3-tts

Zen3 Audio

Collection

Speech recognition + text-to-speech. • 7 items • Updated 3 days ago