MOSS-TTS Nano
Apache-2.0 (code) · weight terms require reviewCompact CPU speech generation with streaming and voice cloning.
Compact CPU speech generation with streaming and voice cloning. Supports 48 kHz stereo through a separate audio tokenizer. Follow the ONNX CPU guide for lightweight deployment; language coverage depends on the checkpoint.
Hear MOSS-TTS Nano
MOSS-TTS Nano 100M · official ONNX CPU checkpoint · synthetic Bella reference · Same three scripts across the directory.
Best for: Compact CPU speech generation with streaming and voice cloning.
Before you choose: Supports 48 kHz stereo through a separate audio tokenizer. Follow the ONNX CPU guide for lightweight deployment; language coverage depends on the checkpoint. The model card does not clearly specify a separate weight license; verify redistribution terms. In this unedited fixed-seed sample, the emotional sentence repeats at the end. Audition your own scripts for repetition.
No comparable hardware benchmark verified. Check the official setup for your device.
Details reviewed 2026-09-14. Hardware figures are estimates unless a benchmark is linked.
Compare with similar
Compare all →Choose an alternative to hear both models together.
Voices
1 voiceThe scripts
Run MOSS-TTS Nano
git clone https://github.com/OpenMOSS/MOSS-TTS-Nano.git
cd MOSS-TTS-Nano
pip install -e .Sources & setup details
MOSS-TTS Nano FAQ
Do I need a GPU to run MOSS-TTS Nano?
MOSS-TTS Nano supports CPU inference. Speed depends on your processor, runtime, and text.
Can MOSS-TTS Nano clone voices?
Yes — MOSS-TTS Nano supports voice cloning from reference audio.
What license is MOSS-TTS Nano released under?
MOSS-TTS Nano is released under the Apache-2.0 (code) · weight terms require review license. Always check the repository for the exact terms — some models license code and weights separately.
What languages does MOSS-TTS Nano support?
MOSS-TTS Nano supports English, Chinese.
Is there a hosted MOSS-TTS Nano API?
OpenSpeech Cloud is accepting production-access interest for MOSS-TTS Nano, but this model is not hosted in the beta. Use the official repository for current deployment options.