Mini toolsHy3
By Tencent
Model family: Hunyuan
Hy3 is Tencent's flagship large language model, built as a 295B parameter Mixture of Experts architecture with 21B activated parameters, an additional 3.8B parameter MTP layer, and a 256K token context window. It follows the Hy3 Preview release, incorporating feedback from over 50 product teams to improve task execution, tool calling stability, and output formatting. The model offers a switchable reasoning mode, using no_think for direct responses or a high reasoning effort setting for deep chain of thought on complex math, coding, and reasoning tasks. Internal evaluations show reduced hallucination rates, fewer commonsense errors, and stronger multi turn context retention across long conversations. Hy3 performs well on agentic workflows, coding benchmarks, and long context tasks, and is released under the Apache License 2.0 with weights available in standard and FP8 quantized formats.
New
Text
Gen 7
Overview
Hy3 is a 295 billion parameter Mixture of Experts language model from Tencent with 21 billion active parameters and a 256K token context window. It supports fast and deep reasoning modes, tool calling, and multi turn dialogue, with improved reliability, reduced hallucination, and stronger agentic and coding performance over its preview version.
About Tencent
Tencent is a world-leading internet and technology company that develops innovative products and services to improve the quality of life of people around the world.
Industry:
Technology, Information and Media
Company Size:
110558
Location:
Shenzhen, CN
Website:
tencent.com
Last updated: July 6, 2026
