Gemini 3.7 Flash began rolling out in GitHub Copilot on August 13, 2026. GitHub says its early testing found improvements over the previous version in web and app development, agentic coding, code quality, codebase research and verification during complex tasks.

Where Gemini 3.7 Flash is available

The model is being added gradually to the model picker in Visual Studio Code, Visual Studio, Copilot CLI, the Copilot cloud agent, the GitHub Copilot app, JetBrains IDEs, Xcode and Eclipse. It is available for Copilot Pro, Pro+, Max, Business and Enterprise plans.

Business and Enterprise administrators must enable the Gemini 3.7 Flash Preview policy before members can select it. GitHub also notes that the model is billed at provider list pricing under Copilot’s usage-based billing.

Why this matters for QA engineers

Repository research and verification are central to agent-assisted QA work. A faster model that more reliably finds relevant code, runs checks and presents a verified result could shorten failure investigation and test-maintenance cycles. However, GitHub’s early results should be treated as a reason to benchmark, not proof that the model will outperform alternatives in every repository.

  • Replay a fixed set of flaky-test, test-generation and regression-fix tasks against the previous model.
  • Measure verified completion, accepted changes, test pass rate, latency and AI-credit cost.
  • Check whether repository research cites the correct files, tests and configuration.
  • Keep human review and CI quality gates in place for generated product or test-code changes.

A practical rollout check

Start with three recent QA tasks: one localized test repair, one repository-wide failure investigation and one multi-step automation change. Run each task more than once because agent results can vary, then compare verified outcomes and cost before changing a team default.

Sources