I remember when people were saying "It's useless to open-source big models because nobody will be ab...
Cerebras 在企业环境中以每秒约1000个token的速度运行千亿参数模型Kimi K2.6,彻底颠覆了‘大模型开源无用’的旧认知,证明硬件创新可使开源大模型实现实用化部署。
入选理由:Cerebras 在企业测试中以约1000 tokens/s的速度运行Kimi K2.6(千亿参数模型),创当前最快推理记录。



![[AINews] Claude Opus 5: Fable-level performance at Opus price (half Fable)](https://substackcdn.com/image/fetch/$s_!1XAa!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F34020777-3ff4-437c-af38-c915ca21b7fd_2464x1352.png)






