Xiaomi's mimo-v2-pro ai model rises to the top, challenging gpt-3 and claude

Xiaomi has made a bold move in the AI landscape, launching its powerful language model MiMo-V2-Pro and positioning it as a formidable competitor to industry giants like GPT-3 and Claude.

Mimo-v2-pro outperforms in benchmarks

Just months after its March debut, MiMo-V2-Pro has exceeded all expectations, outperforming popular models like Claude Sonnet 4.6, Gemini 3 Pro, and GPT-5.2 in extensive benchmark analysis. Xiaomi confirms its AI engine is extremely competitive, even surpassing some benchmarks in preliminary tests.

Affordable and open-source

What sets MiMo-V2-Pro apart is its affordability and open-source nature. While other top models like Claude Sonnet 4.6 and Opus 4.6 come with hefty price tags of $3 to $5 per million tokens of input, MiMo-V2-Pro offers a more accessible pricing structure. Users can opt for a free plan with up to 256,000 tokens of context, or pay $1 per million tokens for up to 1,000,000 tokens of context, with a total cost of between $3 and $6 per million tokens of output.

This cost-effective approach allows Xiaomi to make AI more accessible to developers and businesses, potentially disrupting the market dynamics. In contrast, Anthropic's Claude model limits its subscriptions to only work with their own tools, while Xiaomi is open to third-party integrations with OpenClaw.