September 23, 2026: Claude Opus 5.5 is arriving in GitHub Copilot. GitHub announced the rollout on September 22, adding another model option for agentic coding and longer tasks. See the official release announcement.
Who can use Claude Opus 5.5 in Copilot?
GitHub lists Copilot Pro+, Max, Business and Enterprise as eligible plans. Supported surfaces include VS Code, Copilot CLI, the coding agent, Copilot app, github.com and several other IDEs. Availability is gradual, so the model may not appear for every eligible user immediately.
Business and Enterprise administrators control access through model policies. GitHub says new models are enabled automatically under default model enablement unless an administrator has disabled that default or the individual model. Usage is billed at provider list pricing.
What changed?
GitHub reports that early testing found task resolution comparable to Opus 5 with fewer steps and tokens. That is a vendor-reported result, not a demonstrated improvement in QA defect detection or test reliability.
Why this matters for QA engineers
Our QA assessment: a model rollout is a reason to recheck an assistant workflow before changing team defaults. Use a small set of known failures to assess whether generated fixes preserve meaningful assertions and expose the original defect.
For example, compare the existing model and Opus 5.5 on the same flaky-test investigation. Record the selected model, repository revision, prompt, tool permissions, test results and usage cost. A patch should fail against the known broken behavior and pass after the verified fix; a green run alone can hide weakened coverage.
Teams should also confirm model availability and administrator policy before scheduling an evaluation. Keep adoption decisions tied to reproducible results rather than assuming fewer agent steps mean more reliable tests.
Source
GitHub Changelog: Claude Opus 5.5 is now available in GitHub Copilot — September 22, 2026; verified September 23, 2026.