Updated

ZONOS3-Mini

Zyphra released ZONOS3-Mini, an open-weight multilingual text-to-speech model with voice cloning and fine-grained vocal controls. Zyphra says the model supports 43 languages and can run locally in browsers and native environments.

By No Code MBA · Facts checked

Provider
Zyphra
Announcement date
Status at last check
Generally available

What changed

ZONOS3-Mini is Zyphra's new compact TTS release, using a 15 Hz dual audio codec and a sparse mixture-of-experts backbone with about 100 million active parameters. Zyphra says it adds real-time local deployment, multilingual synthesis, voice cloning, and controllable delivery.

Access and availability

Download Apache 2.0 model weights from Zyphra's Hugging Face repository, or use its browser and native runtimes. Zyphra also offers hosted access through Zyphra Cloud.

Official access documentation

What you can build

Test voice cloning and language quality with consented reference audio, and check pronunciation and output controls before using generated speech in a production workflow.

Limitations

The model generates speech rather than general text responses, and its local browser runtime requires WebGPU support. Zyphra's performance and quality measurements are company-reported.

Sources and release history