腾讯混元把295B的旗舰模型量化到1-bit和4-bit，单卡就能跑，还免费两周API，适合做智能体。
腾讯混元发布了Hy3模型的1-bit和4-bit量化版本，原始模型为295B参数的MoE架构，在同等规模中性能最佳，可媲美万亿参数旗舰模型。量化后模型可在单张GPU上运行，支持llama.cpp和MTP（多令牌预测）。模型采用Apache 2.0协议，对商用友好，并提供两周免费API试用。
We’ve just released the 1-bit & 4-bit version of Hy3, a flagship-scale 295B model that can be served...
We’ve just released the 1-bit & 4-bit version of Hy3, a flagship-scale 295B model that can be served on a single GPU. 👌 Run Hy3 with llama.cpp, enable MTP, and experience powerful intelligence on dramatically lower hardware.🚀🚀🚀 Can’t wait to see what you buil #Hy3 # #Hy #GGUF G #llamacpp acpp Tencent Hy @TencentHunyuan 🚀Hy3 is here. 295B MoE. Best in its size class. Rivals trillion-scale flagships. Reliable and affordable for most agentic usecases. Apache 2.0. Friendly for commercial use. FREE API for 2 weeks → openrouter.ai/tencent/hy3:fr… i � huggingface.co/tencent/Hy3 SL hy.tencent.com/research/hy3 cW4 🔗 View Quoted Tweet 💬 3 🔄 12 ❤️ 53 👀 1906 📊 13 ⚡