罗福莉加入小米后首秀,解释 MiMo-V2-Flash 模型如何做到推理速度飞快

IT之家12-17

   IT之家12月17日消息,2025小米人车家全生态合作伙伴大会于今日举行,Xiaomi MiMo大模型负责人罗福莉迎来入职后首秀。

  小米昨日晚间惊喜发布了Xiaomi MiMo-V2-Flash开源MoE模型,总参数量309B,活跃参数量15B,专为智能体AI设计,专注于快。不少IT之家小伙伴体验后发现,该模型推理速度非常快。

  罗福莉解释称,MiMo-V2-Flash围绕极致推理效率设计了模型结构,通过3层MTP推理加速并行Token验证,实现了2.0~2.6倍的推理速度提升。

   MiMo-V2-Flash凭借总参数309B(激活15B),实现了代码和Agent评测基准上全球开源模型Top2,且初步具备模拟世界的能力,可通过HTML写操作系统、模拟太阳系、画一棵圣诞树等。

  罗福莉还谈到了下一代智能体系统,认为下一代智能体系统不是一个“语言模拟器”,而是一个真正能够理解世界、并与之共存的“智能体”。

  2025小米“人车家全生态”合作伙伴大会专题

关注同花顺财经(ths518),获取更多机会

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Comments

We need your insight to fill this gap
Leave a comment