Microsoft Prioritizes Anthropic’s Claude Sonnet 4 Over OpenAI Models with New Auto AI Model Selection in Visual Studio Code

Microsoft is rolling out a new auto model selection feature in its Visual Studio Code (VS Code) editor, marking a strategic move to optimize AI-powered coding assistance by primarily favoring Anthropic’s Claude Sonnet 4 model over OpenAI’s GPT-5 variants. This feature, now in preview, automatically selects the most suitable AI model for developers’ requests within VS Code’s GitHub Copilot environment, aiming to deliver faster response times, reduced rate limiting, and cost savings for paid users.

The auto model selection feature dynamically picks the best AI language model based on real-time availability, capacity, and performance. It currently toggles between Anthropic’s Claude Sonnet 4, OpenAI’s GPT-5, GPT-5 mini, and other integrated models, unless specific models have been disabled by an organization’s policy. For paid Copilot users, Claude Sonnet 4 is the primary model powering this automatic selection, underscoring Microsoft’s emphasis on Anthropic’s technology as central to their AI coding assistant’s enhanced performance.

Once a model is selected at the beginning of a chat session, the system uses the same model throughout that interaction, although Microsoft plans to evolve this approach to tweak model assignments based on task complexity in future updates. This means that developers enjoy consistently optimized AI code completions and chat assistance, without needing to manually switch between models and without facing interruptions from rate limits.

From a cost perspective, Microsoft offers a 10% discount on premium AI requests for paid users leveraging the auto model selection. This is coupled with a clever multiplier system that accounts for the relative “cost” of each model. For instance, Claude Sonnet 4 requests count as 0.9 times a premium request, GPT-5 mini counts as 0x (completely free for paid users), and if premium request quotas are exceeded, auto will default to a 0x multiplier model, ensuring uninterrupted usage. This pricing strategy not only incentivizes using the Anthropic-powered model but also manages operational costs while maintaining high-quality AI assistance.

This development positions Microsoft’s VS Code as one of the most advanced code editors for AI-assisted programming. Alongside existing tools like Microsoft IntelliCode—which focuses on local, context-aware code completions—and AWS CodeWhisperer integration, VS Code is enhancing its AI ecosystem by offering greater flexibility and smarter resource allocation among various AI models.

Microsoft’s strategic favoring of Anthropic’s Claude Sonnet 4 aligns with broader industry trends where multiple AI providers compete to deliver the most efficient, accurate, and developer-friendly AI solutions. Anthropic, known for its focus on AI safety and alignment, appears to have gained trust from Microsoft as a preferred backend model provider, supplementing but not replacing OpenAI’s offerings in VS Code’s AI stack.

In addition to model selection improvements, Microsoft continues to invest in rich AI workflows within VS Code through initiatives like the AI Toolkit for Visual Studio Code, enabling developers to download, fine-tune, and deploy AI models from within the editor itself. This holistic approach empowers developers to harness AI not just for code completion but for building AI-powered applications seamlessly within their development environment.

Overall, this latest update demonstrates Microsoft’s commitment to integrating diverse AI technologies while optimizing cost, performance, and user experience in VS Code’s AI assistant capabilities. By automating model selection and preferentially leaning on Anthropic’s Claude Sonnet 4, VS Code users gain faster, smarter coding help tailored dynamically to their workload, marking a new chapter in AI-augmented software development tools.