Microsoft AI Models Outperform Claude & Google Nano Banana

Microsoft has unveiled seven new AI models, claiming they outperform competitors in blind evaluations and benchmarks. The company states its MAI-Thinking-1 model surpassed Anthropic's Claude Sonnet 4.6 in blind tests and matched Claude Opus 4.6 on a leading coding benchmark. What's interesting is that Microsoft also says its MAI-Image-2.5 models beat Google's Nano Banana 2 on image-editing leaderboards. This launch marks Microsoft's most ambitious effort to develop its own frontier AI models, alongside its partnership with OpenAI. The bottom line is that this intensifies competition among leading AI developers, potentially bringing new capabilities to users.