Moonshot AI 在 Hugging Face 开放 Kimi K3.1 Pro 模型权重
来源标题：Moonshot AI releases Kimi K3.1 Pro model weights
Moonshot AI 在 Hugging Face 上公布了旗舰多模态模型 Kimi K3.1 Pro 的权重，可供独立部署。该模型采用混合专家架构，总参数 2.8 万亿，每次推理约激活 1040 亿，上下文窗口约 100 万 tokens。华为同时宣布通过 vLLM Ascend 和 SGLang 为其 Atlas A3 与 Ascend 950 系统提供推理支持。
以下正文同步自 动点科技 TechNode，版权归原站所有，已转换为易读排版。
Moonshot AI has released the weights of Kimi K3.1 Pro on Hugging Face, making its flagship multimodal model available for independent deployment. The model uses a mixture-of-experts architecture with 2.8 trillion total parameters and approximately 104 billion activated per inference, and supports a context window of around one million tokens. K3.1 Pro builds on Kimi K3 with expanded reinforcement learning for coding, 3D creation, financial research and long-running agent tasks. The weights release follows the model’s earlier availability through Kimi’s applications and API. Huawei has also announced inference support through vLLM Ascend and SGLang, covering its Atlas A3 and Ascend 950 systems. [Moonshot AI]
正文由 探How 从来源站点 RSS 同步，内容未经改写；遇到排版缺失或需要图片、视频时请以原文为准。