Anthropic's latest Claude model, Claude Sonnet 5.5, is now generally available in GitHub Copilot. The model targets well-scoped everyday development tasks like building features and fixing bugs, and early testing shows it matches Claude Sonnet 5 on coding performance while completing tasks faster and using fewer tokens and tool calls.
The model operates under usage-based billing at provider list pricing. It's available to Copilot Pro, Pro+, Max, Business, and Enterprise users across a wide range of environments, including Visual Studio Code, Visual Studio, JetBrains IDEs, Xcode, Eclipse, Copilot CLI, the GitHub Copilot coding agent, the GitHub Copilot app, github.com, and GitHub Mobile on iOS and Android. Rollout is gradual, so some users may not see it immediately.
What early testing revealed
GitHub's internal evaluation highlighted Sonnet 5.5's efficiency gains. The model delivered coding results comparable to Claude Sonnet 5 while consuming significantly fewer steps, tokens, and tool calls. It also finished tasks noticeably faster - a practical advantage for developers who need quick, accurate responses during routine work.
This efficiency profile matters in usage-based billing scenarios. Fewer tokens per task can translate directly to lower costs for teams and individual developers who rely on Copilot throughout their workday.
Administrator controls for Business and Enterprise
Copilot Enterprise and Copilot Business plan administrators can manage access to Claude Sonnet 5.5 through the model policy settings in Copilot. Under the default configuration, new models are automatically enabled. Administrators can restrict access by turning off the global default or explicitly disabling this specific model.
This gives organizations control over which AI models their developers can use, allowing teams to evaluate Sonnet 5.5 on their own codebases before broader rollout.
Where to find it and give feedback
Developers can select Claude Sonnet 5.5 from the model picker in any supported environment once it appears in their account. GitHub encourages users to share feedback through the GitHub Community. For those looking to build skills with AI-assisted development, AI for Software Developers Courses offer structured learning paths that cover integrating tools like Copilot into daily workflows.
Why this matters for IT and development professionals
Claude Sonnet 5.5's combination of maintained coding accuracy and reduced token consumption means faster, cheaper completions for everyday tasks. For developers working under usage-based Copilot billing, that efficiency gain is tangible: same quality output, lower per-task cost. Teams evaluating which models to enable should test Sonnet 5.5 against their typical bug fixes and feature work - the speed difference alone could shift daily workflows.
Your membership also unlocks:
