whatmodel.sh / tts / ZipVoice
ZipVoice
↗ share
+ contribute
ZipVoice is an open-weights text-to-speech model from k2-fsa (Xiaomi), released June 2025. Its published size is 0.123B parameters. Licence: apache-2.0.
Catalogue identity notes
k2-fsa/ZipVoice model 'zipvoice' (123M parameters per README), not zipvoice_distill or ZipVoice-Dialog. Licence from the GitHub repository; the HF card states none.
Score within the Vapi Humanness · Bradley–Terry rating, 2026-09-25 snapshot comparison set only.
$ how it compares
The text-to-speech models ranked either side of ZipVoice, out of 148 with a score, then more from k2-fsa (Xiaomi).
| # | model | score | where | ||
|---|---|---|---|---|---|
| 40 |  F5-TTS v1 Base | 62±3 | local | ||
| 41 |  Inworld TTS-1.5 Max | 62±3 | api $10.00/M chars | ||
| 42 | ZipVoice | 62±3 | local | ||
| 43 | Ghana TTS 72k | 62±7 | comparison only | ||
| 44 |  Qwen3-TTS 12Hz 0.6B Base | 62±2 | local | ||
| more from k2-fsa (Xiaomi) | |||||
| 18 | OmniVoice | 66±2 | local | ||
see all text-to-speech models · find the ones that fit your machine
$ what it's good at
voice cloning▁·▁▁▁▁▁▁▇█▃▃ #18 of 37 · 80% of its score · 4 tests
Beats Gemini 3.1 Flash TTS Preview, Fun-CosyVoice3-0.5B-2512, MOSS-TTS v1.5 (8B) here, though they rank higher overall. Trails Qwen3-TTS 12Hz 0.6B Base here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Seed-TTS test-en · speaker similarity (WavLM-large SV) | 69.7% ✎ | ▁·▂▂▂█▆▃▅··▂ beats 12/30 | 32% |
| Seed-TTS test-zh · speaker similarity (WavLM-large SV) | 75.1% ✎ | ▁·▂··▂·█▄·▃▄ beats 6/25 | 27% |
| Seed-TTS test-en · Whisper large-v3 WER | 1.7% ✎ | ▁···▂▂▂▂█▆▄▂ beats 22/35 | 12% |
| Seed-TTS test-zh · Paraformer CER | 1.4% ✎ | ▁·▂▁▁▁▂▃▅█·▂ beats 9/30 | 8.1% |
other tests2 not in a task · 16% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| RVCBench LibriTTS clean voice cloning · ECAPA speaker similarity | 58.0% ✎ | ▃·····▂▂▂▃█▆ beats 12/15 | 16% |
| RVCBench LibriTTS clean voice cloning · Whisper-medium WER | 5.0% ✎ | ▂▂▂··▂·▃▃▃█▂ beats 12/15 | 0.2% |
$ where to run it
Other ways to run it
version
- Run it as published · needs ~0.55 GB · from github.com/k2-fsa (official)
git clone https://github.com/k2-fsa/ZipVoice.git && cd ZipVoice && pip install -r requirements.txt && python3 -m zipvoice.bin.infer_zipvoice --model-name zipvoice --prompt-wav prompt.wav --prompt-text "I am the transcription of the prompt wav." --text "I am the text to be synthesized." --res-wav-path result.wav
Official weights and documentation: huggingface.co/k2-fsa ↗
$ how the score was worked out · OmniVoice: Towards Omnilingua… 80%, RVCBench: Benchmarking Robust… 16%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 OmniVoice: Towards Omnilingual Zero-Shot Text-to-Speech with Diffusion Language Models (arXiv:2604.00688v3) | 4 | 80% |
| RVCBench: Benchmarking Robustness of Voice Cloning Across Modern Audio Generation Models (arXiv:2602.00443v2) | 2 | 16% |
| prior · pull toward the middle | 5% | |
| weighted average | = 61.9 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json