Xiaomi

Xiaomi's frontier omni-modal model that natively processes image, video, and audio inputs within a unified architecture. Combines strong multimodal perception with agentic capability including visual grounding, multi-step planning, tool use, and code execution.

Xiaomi's flagship foundation model with over 1 trillion parameters and 1M context length. Deeply optimized for agentic scenarios. Ranks among global top tier on standard benchmarks, approaching Opus 4.6 levels.