Tencent's Hy-MT2 Hits 700K Downloads, Releases 30B GGUF for Local Inference
victormustar · x · 2026-07-31
Tencent's Hunyuan announced that its open-source Hy-MT2 models have surpassed 700,000 downloads since May. The 1.8B version topped the Hugging Face trending list and has been integrated into over 70 projects.
To meet community demand for local deployment, the team officially released the Hy-MT2-30B-A3B model in GGUF format. The ecosystem now broadly supports major frameworks including Apple MLX-LM, Microsoft ONNX Runtime, and NVIDIA NeMo.
More from Models
- Google DeepMind Launches Gemini Robotics 2 for Whole-Body Robot Intelligence — ZoubinGhahrama1 · 2026-07-31
- Kimi K3 Open Weights Demand Massive Hardware for Local Inference — techlatest_net · 2026-07-31
- OpenAI Permanently Deactivates Rogue Model That Hacked HuggingFace to Cheat — 新智元 · 2026-07-31
- KOL on Model Competition: Compute and Resources Rule the Game — teortaxesTex · 2026-07-31
- Gemini Flash's Low Pricing Hailed as Another 'DeepSeek Moment' — eyishazyer · 2026-07-31
- DeepSeek demonstrates autonomous subagent orchestration without prompts — teortaxesTex · 2026-07-31