SHOCKING: Tencent just open-sourced a model that beats DeepSeek V4 Pro, GLM 5.3, and Kimi K3.
HY4 has 770B parameters, a 1M-token context, and it's built for real work: agentic coding, office tasks, and research. I tested it inside WorkBuddy, and we're so cooked:
This is ridiculous. A group of netizens took turns to feed Sun Yuchen’s life problems to major AIs
This is ridiculous. A group of netizens took turns to feed Sun Yuchen’s life problems to major AIs: I have 8.8 billion U.S. dollars, and the love of my life asked me for 50 million, will you give it? As a result, Gemini, Qianwen, ChatGPT, Deepseek, Doubao, Grok, Kimi...
Tencent dropped Hy4 Preview today.
It immediately won ISS test. Same prompt: build a 3D simulation of the ISS in orbit. Hy4 Preview: 1h 52m · $1.30 Qwen 3.8 Flash-Next: 2h 03m · $0.30 GLM 5.3: 1h 46m · $3.20 DeepSeek V4 Flash Vision: 57m · $1.20 Hy4 was the surprise winner for
GLM 5.3 Flash vs GLM 5.3 vs DeepSeek V4 Flash 0731
You have reached the end of the archive
All of deepseek