Microsoft is implementing automatic AI model selection within its Visual Studio Code editor. this feature aims to optimize performance by automatically choosing the most suitable AI model. the auto model feature will utilize models like Claude Sonnet 4, GPT-5, and GPT-5 mini for GitHub Copilot free users. However,those with paid subscriptions will “primarily rely on Claude Sonnet 4.”

this move signals a preference by microsoft for Anthropic’s AI models over OpenAI’s GPT-5 models for coding and development tasks. Sources indicate that Microsoft has been directing its developers to utilize Claude Sonnet 4 in recent months.

According to an internal email from june, “Based on internal benchmarks, Claude Sonnet 4 is our recommended model for GitHub Copilot,” stated Julia Liuson, head of microsoft’s developer division. this guidance remains in effect even after the release of GPT-5.

Microsoft is also making “notable investments” in training its own AI models. Microsoft AI chief Mustafa Suleyman mentioned in a recent employee town hall that, “We’re also going to be making significant investments in our own cluster. So today, MAI-1-preview was only trained on 15,000 H100s, a tiny cluster in the grand scheme of things.”

Moreover, Microsoft intends to integrate Anthropic’s AI models into certain Microsoft 365 applications. The Facts reports that Microsoft 365 Copilot will be “partly powered by Anthropic models,” following findings that these models outperformed OpenAI in Excel and powerpoint.

OpenAI and Microsoft announced a new deal last week, potentially paving the way for the AI startup’s initial public offering. Microsoft has invested over $13 billion in OpenAI since 2019 and maintains a