---
title: "腾讯发布Hy3：295B参数MoE模型，21B激活参数，Apache 2.0许可"
source: "Simon Willison 博客（Atom）"
source_tier: "t2"
url: "https://simonwillison.net/2026/Jul/6/hy3/#atom-everything"
permalink: "https://feed.astralor.com/items/itm_27f14f334409a8ad7749b784"
published: "2026-07-06T23:57:35.000Z"
score: 75
channel: "ai"
tags: ["Hy3","Mixture-of-Experts","开源大模型","腾讯"]
---

# 腾讯发布Hy3：295B参数MoE模型，21B激活参数，Apache 2.0许可

## AI 点评

> 这个模型的亮点是‘以小博大’——仅21B激活参数就敢叫板70B+模型，说明MoE路线在效率上确实厉害。对预算有限又想跑大模型的中小团队简直是福音。不过目前还没有看到独立评测，建议亲自在OpenRouter上跑几个任务验证一下。

## 速览

腾讯Hy团队推出Hy3模型，295B参数MoE架构，仅21B激活参数，3.8B MTP层参数。基于4月预览版反馈，利用更高质量数据进行后训练，性能超越同规模模型，媲美2-5倍参数的开源旗舰模型。上下文长度256K，模型权重598GB（FP8量化版300GB）。即日起至7月21日在OpenRouter免费使用。

## 关键点

- 参数规模：295B总参数，21B激活参数，3.8B MTP层参数，MoE架构。
- 性能表现：超越同规模模型，媲美2-5倍参数的开源旗舰模型。
- 上下文长度：256K tokens。
- 开放许可：Apache 2.0，免费使用至7月21日（OpenRouter）。

## 全文

tencent/Hy3 New Apache 2.0 licensed model from Tencent in China: Hy3 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+ products and scaled up post-training with higher quality data. Today, we introduce Hy3, which outperforms similar-size models and rivals flagship open-source models with 2-5x parameters. It also shows significant gains in utility across various products and productivity tasks. The full-sized model is 598GB on Hugging Face, and the FP8 quantized one is 300GB. The context length is 256K. It's available for free on OpenRouter until July 21st. I had it "Generate an SVG of a pelican riding a bicycle" there and got this: Tags: ai, generative-ai, llms, pelican-riding-a-bicycle, llm-release, ai-in-china

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_27f14f334409a8ad7749b784