Tencent Debuts Hunyuan 3.0 with Major Programming Boost
Tencent has recently unveiled its latest-generation AI large model, Huyuan 3.0, named Hy3, marking a significant step forward in the company's artificial intelligence efforts, particularly in programming capabilities. With the addition of AI expert Yao Shunyu, previously recruited from a major internet company, the model's performance has seen notable improvements, sparking fresh expectations across the industry.
Currently available for free under the name Hy3preview, users can test it on OpenRouter. Compared to the previous Huyuan 2.0 model, Hy3 has made a substantial leap in programming. According to user-provided SWE-Bench programming test results, Huyuan 2.0 scored only 53.0%, while Huyuan 3.0 reached 74.4%—a performance gain of over 40%. This progress brings Hy3's programming abilities close to GLM-4.7, and though it still lags behind GLM-5, it undoubtedly ranks among the top domestic AI large models.

Huyuan 3.0 adopts a Mixture of Experts (MoE) architecture with three levels of reasoning, handling context up to 262K tokens and achieving a speed of 23 tokens per second. This multimodal enhancement gives Hy3 an edge in managing complex tasks.
Moreover, the release of Huyuan 3.0 comes amid intensifying competition in the AI field. Tencent's new model not only directly competes with rivals like DeepSeek V4 but also opens up fresh possibilities for future AI development. As artificial intelligence technology continues to evolve, Tencent aims to carve out a market position with Huyuan 3.0 and challenge industry leaders such as OpenAI.

In summary, the launch of Huyuan 3.0 is not only a key move by Tencent in AI but also provides a more powerful tool for a wide range of developers and users. Its performance in future applications is highly anticipated.
Related article
Claude Opus 5.2 Night Gray Launches with Faster Response, Solving Laziness Issue
Opus 5.2 quietly launched this morning, prompting many developers to notice that the updated model, Claude Opus 5.2, has begun a limited rollout within Claude Code.Yesterday evening, X users observed that invoking Opus 5 in Claude Code yielded perfor
Fireworks AI Unveils FireRouter with Opus: Cuts Encoding Costs by 57% with Minimal Accuracy Drop
Fireworks AI has introduced FireRouter with Opus, the industry’s first cache-aware routing system tailored for the Claude Opus series. Now accessible via a serverless endpoint, this independent routing model underwent over a month of internal A/B tes
NVIDIA Unveils Nemotron-Labs-Audex-30B-A3B Unified Audio Intelligence Model
As multimodal large models evolve rapidly, audio processing capabilities are frequently compromised—many models improve audio understanding at the expense of text logic. To address this, NVIDIA researchers have introduced Nemotron-Labs-Audex-30B-A3B
Related Special Topic Recommendations
Comments (1)
0/500
Tencent has recently unveiled its latest-generation AI large model, Huyuan 3.0, named Hy3, marking a significant step forward in the company's artificial intelligence efforts, particularly in programming capabilities. With the addition of AI expert Yao Shunyu, previously recruited from a major internet company, the model's performance has seen notable improvements, sparking fresh expectations across the industry.
Currently available for free under the name Hy3preview, users can test it on OpenRouter. Compared to the previous Huyuan 2.0 model, Hy3 has made a substantial leap in programming. According to user-provided SWE-Bench programming test results, Huyuan 2.0 scored only 53.0%, while Huyuan 3.0 reached 74.4%—a performance gain of over 40%. This progress brings Hy3's programming abilities close to GLM-4.7, and though it still lags behind GLM-5, it undoubtedly ranks among the top domestic AI large models.

Huyuan 3.0 adopts a Mixture of Experts (MoE) architecture with three levels of reasoning, handling context up to 262K tokens and achieving a speed of 23 tokens per second. This multimodal enhancement gives Hy3 an edge in managing complex tasks.
Moreover, the release of Huyuan 3.0 comes amid intensifying competition in the AI field. Tencent's new model not only directly competes with rivals like DeepSeek V4 but also opens up fresh possibilities for future AI development. As artificial intelligence technology continues to evolve, Tencent aims to carve out a market position with Huyuan 3.0 and challenge industry leaders such as OpenAI.

In summary, the launch of Huyuan 3.0 is not only a key move by Tencent in AI but also provides a more powerful tool for a wide range of developers and users. Its performance in future applications is highly anticipated.
Claude Opus 5.2 Night Gray Launches with Faster Response, Solving Laziness Issue
Opus 5.2 quietly launched this morning, prompting many developers to notice that the updated model, Claude Opus 5.2, has begun a limited rollout within Claude Code.Yesterday evening, X users observed that invoking Opus 5 in Claude Code yielded perfor
Fireworks AI Unveils FireRouter with Opus: Cuts Encoding Costs by 57% with Minimal Accuracy Drop
Fireworks AI has introduced FireRouter with Opus, the industry’s first cache-aware routing system tailored for the Claude Opus series. Now accessible via a serverless endpoint, this independent routing model underwent over a month of internal A/B tes
NVIDIA Unveils Nemotron-Labs-Audex-30B-A3B Unified Audio Intelligence Model
As multimodal large models evolve rapidly, audio processing capabilities are frequently compromised—many models improve audio understanding at the expense of text logic. To address this, NVIDIA researchers have introduced Nemotron-Labs-Audex-30B-A3B





Home






