Xiaomi Releases 1-Trillion-Parameter AI Model That Benchmarks Close to GPT-5.2 and Claude Opus 4.6

Xiaomi released MiMo V2 Pro, a one-trillion-parameter AI model that benchmarks within striking distance of OpenAI's GPT-5.2 and Anthropic's Claude Opus 4.6, according to VentureBeat. The model represents a significant leap in capability for a Chinese hardware company better known for smartphones and consumer electronics. MiMo V2 Pro is open-weight, allowing developers to fine-tune and deploy it independently. The release intensifies competition in the frontier model space, where a growing number of companies outside the traditional US AI labs are producing competitive systems.