Tencent Releases Hy3: A 295B-Parameter MoE AI Model with 21B Active Parameters
Summary
Tencent's Hy Team has released Hy3, a 295B-parameter Mixture-of-Experts (MoE) AI model with 21B active parameters and 3.8B MTP layer parameters. Building on feedback from 50+ product teams after the Hy3 Preview launch in late April, the team fixed task execution and interaction issues while improving post-training pipeline quality and scale. The model significantly outperforms similar-sized models.
Source
Twitter / XTencent Releases Hy3: A 295B-Parameter MoE AI Model with 21B Active Parametershuggingface.coKey quotes
· 3 pulledHy3 is a 295B-parameter Mixture-of-Experts (MoE) model with 21B active parameters and 3.8B MTP layer parameters, developed by the Tencent Hy Team.
Following the Hy3 Preview launch in late April, we gathered feedback from 50+ product teams.
We fixed various issues in task execution and interaction, and improved both the quality and scale of our post-training pipeline.
You might also wanna read
Tencent ships Hy3 as an Apache 2.0 agent model
Tencent Launches Hunyuan Hy3 LLM with Strong Enterprise Performance
Tencent's Apache-licensed Hy3 takes on GLM-5.2 at half the size — and wins everywhere except coding
Tencent's Apache-licensed Hy3 takes on GLM-5.2 at half the size — and wins everywhere except coding
DeepSeek-V4 Series Preview: Million-Token Context MoE Models with 1.6T Parameters
DeepSeek introduces the V4 series of Mixture-of-Experts (MoE) language models, including DeepSeek-V4-Pro (1.6T parameters, 49B activated) an
Tencent Releases Open-Source Hunyuan3D-2.1 AI for Generating High-Fidelity 3D Assets from Images
Tencent has released Hunyuan3D-2.1, a fully open-source AI model for generating high-fidelity 3D assets with physically-based rendering (PBR
Product Hunt·1y ago
GLM-4.7-Flash: Z.ai's 30B-A3B MoE Model for Lightweight AI Deployment
GLM-4.7-Flash is a 30B-A3B Mixture of Experts (MoE) model developed by Z.ai, positioned as the strongest model in the 30B parameter class. T
Comments
Sign in to join the conversation.
No comments yet. Be the first.