Microsoft debuts MAI-Thinking-1 AI model as it reduces reliance on OpenAI technology

Microsoft introduced MAI-Thinking-1, a midsize AI model intended for high-efficiency, low-token-cost situations. Microsoft claimed the model matches Anthropic's Claude Opus 4.6 on the SWE-Bench Pro coding benchmark, and independent testers preferred it to Anthropic's Sonnet 4.6.

The move is part of Microsoft's shift toward a model-agnostic approach that lets users select which AI models to use, including options from Anthropic. This reflects Microsoft's effort to decrease dependence on OpenAI's technology for its Copilot offerings.

Microsoft has poured roughly $13 billion into OpenAI over the years. However, the relationship has deteriorated: Microsoft lost its status as OpenAI's exclusive cloud provider in January 2025, lost its right of first refusal in October, and in April lost exclusive access to OpenAI's intellectual property, though it no longer has to make revenue-sharing payments.