AI Models
Tencent Open-Sources Hy3, a 295B Model That Beats GLM-5.2 at Half the Size
Hunyuan 3.0 is a 295-billion-parameter mixture-of-experts model that fires just 21B parameters per token, ships under Apache 2.0, and edges GPT-5.5 on science — free on OpenRouter through July 21.
2d ago
