Xiaomi Drops 1T-Parameter AI Model That Rivals GPT-5.2
Xiaomi's MiMo-V2-Pro, codenamed Hunter Alpha, benchmarks near GPT-5.2 and Opus 4.6 in a bold frontier AI play.
Xiaomi just entered the heavyweight ring. The company released MiMo-V2-Pro, a 1-trillion-parameter foundation model codenamed Hunter Alpha that it claims performs close to GPT-5.2 and Opus 4.6 on benchmarks.
The project is led by Fuli Luo, who previously worked on DeepSeek R1 — the model that shook up the AI landscape. Luo describes the release as a "quiet ambush" on the global frontier of AI development.
That's fighting words. Xiaomi, best known for smartphones and consumer hardware, is now gunning directly for the top tier of foundation model makers. A trillion parameters puts MiMo-V2-Pro in rare company, and if the benchmark claims hold up under independent testing, it signals a serious new competitor from China's tech ecosystem.
The smartphone giant clearly isn't content staying in its lane anymore.