ZONOS3-Mini 发布：886M 参数零样本语音克隆 TTS 可实时跑在手机上
ZONOS3-Mini Technical Report
Zyphra 开源了这个 886M 参数的语音克隆模型，100M 激活量就能在手机上实时跑，克隆效果还能对标更大模型。做端侧语音应用、多语言 TTS 的开发者可以直接拿权重和训练代码上手试。
Zyphra 发布 ZONOS3-Mini 技术报告，这是一款超紧凑、可零样本声音克隆的 TTS 模型，总参数量 886M，每个 token 仅激活 100M，可在笔记本和智能手机上实时运行。
技术核心有两点：一是采用 latent MoE 主干网络降低单 token 计算量；二是配套自研音频编解码器 ZAC-1，能以 15 Hz 帧率表示 44.1 kHz 音频。在新基准 ZTTS2-Eval 上，模型覆盖 55 个语言集（35 个干净语音、20 个真实场景），macro-average WER 为 6.71%，与更大、更慢的主流 SOTA 模型相当。
模型权重、推理和训练代码均已开源，对需要端侧语音合成和声音克隆的开发者来说是可直接落地的选择。
ZONOS3-Mini Technical Report
ZONOS3-Mini is an ultra-compact, ultra-fast, zero-shot voice-cloning TTS model. The model can run in real time on a laptop or smartphone. For its latency and size, it achieves state-of-the-art accuracy and cloning fidelity. To achieve this we propose a latent MoE backbone with only 886M parameters (100M active per token) and a new audio codec, Zyphra Audio Codec 1 (ZAC-1), which represents 44.1 kHz audio at a 15 Hz frame rate. On the new ZTTS2-Eval benchmark, ZONOS3-Mini reaches a 6.71% macro-average WER over 55 language sets (35 clean and 20 in the wild), which is competitive with larger and slower state-of-the-art models. We release the model weights, inference, and training code.
This report is categorized under AITOP's "Models" desk. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include ZONOS3-Mini, TTS, 语音克隆, 开源模型.
Source attribution: curated from arXiv cs.AI (@Gabriel Clark, Sofian Mejjoute, George Close, Beren Millidge).
