Zyphra has launched ZONOS3-Mini, an open-source text-to-speech model designed for real-time, on-device performance. This model features a sparse Mixture of Experts architecture with 800 million total parameters and 100 million active parameters, enabling zero-shot voice cloning and fine-grained control over audio delivery. ZONOS3-Mini is optimized to run across various platforms, including web browsers, laptops, and servers, with support for Nvidia GPUs, Apple Silicon, and general CPUs. AI
IMPACT This release offers a high-quality, efficient open-source TTS model that can run on consumer hardware, potentially lowering barriers for real-time voice applications.
RANK_REASON Release of an open-source model with detailed technical specifications and benchmark results. [lever_c_demoted from research: ic=1 ai=1.0]
- Apache Software License 2.0
- Apple Silicon
- M3 MacBook Air
- Nvidia
- NVIDIA H100
- Qwen3-TTS
- ZAC-1
- ZONOS2
- ZONOS3-Mini
- Zyphra
- Zyphra Audio Codec 1
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →