release aidev NEEDS REVIEW SIG 1/5

Xiaomi publishes MiMo V2.5 Base multimodal model

Xiaomi published MiMo V2.5 Base, a multimodal model supporting vision, language, audio, and video understanding with long-context capabilities, hosted on Hugging Face under the MIT license.

PUBLISHED2026-04-27
OBSERVED2026-08-11
AGE4mo
SOURCES1
  • Repository: XiaomiMiMo/MiMo-V2.5-Base
  • Pipeline: text-generation
  • Library: transformers
  • Format: safetensors, fp8
  • Capabilities (from tags): multimodal, vision-language, audio, video-understanding, long-context, conversational, agent
  • Languages: en, zh
  • License: MIT
featuremodel-supportcapability

COMMUNITY

No curated reactions recorded for this event. Facts and takes are kept in separate layers — community context is added by hand, never blended into the record above.