Microsoft has begun rolling out MAI-Code-1.1-Flash, its newest small-tier coding model, inside GitHub Copilot, giving developers a cheaper option that also reads images. The model adds native vision support for image understanding and improves coding quality, instruction following, tool use, and overall performance over the 1.0 release it replaces, according to GitHub’s official changelog MAI-Code-1.1-Flash available in GitHub Copilot.
The biggest draw for teams watching their AI bills is price. GitHub says continued gains in model and serving efficiency let it list MAI-Code-1.1-Flash at 73% lower list price than MAI-Code-1-Flash, the June launch model MAI-Code-1.1-Flash available in GitHub Copilot. Annual Copilot subscribers pay a 0.25× premium request multiplier when the model is used.
Microsoft’s own writeup frames the efficiency gains more concretely. Compared with the model it shipped at Build in June, 1.1 produces higher-quality code at 25% greater token efficiency and a quarter of the cost MAI-Code-1.1-Flash: Better, faster, at a quarter of the cost. In production inside Copilot, Microsoft measured a 22% improvement on Terminal-Bench 2.1 for CLI tasks and a 15% improvement on .NET tasks MAI-Code-1.1-Flash: Better, faster, at a quarter of the cost.
Microsoft AI also reports engagement signals it tracks in production, and these are vendor-measured figures rather than independent tests. Microsoft AI says code survival—how often generated code stayed in a codebase—rose 4% and return visits increased 9% MAI-Code-1.1-Flash: Better, faster, at a quarter of the cost, gains it attributes to the same efficiency work GitHub credits for the lower list price MAI-Code-1.1-Flash available in GitHub Copilot. By Microsoft AI’s account, tokens stream 25% faster and the model needs 25% fewer tokens to finish a task after optimization across hundreds of thousands of reinforcement-learning environments inside Copilot MAI-Code-1.1-Flash: Better, faster, at a quarter of the cost.
Rollout differs by plan. Copilot Free and Student users get 1.1-Flash through auto model selection, while Pro, Pro+, Max, Business, and Enterprise subscribers can pick it manually in the model picker across Copilot CLI, the cloud agent, the Copilot app, Chat, VS Code, Visual Studio, GitHub Mobile, JetBrains, Eclipse, and Xcode MAI-Code-1.1-Flash available in GitHub Copilot. Enterprise and Business admins must switch on the MAI-Code-1.1-Flash policy in Copilot settings, which is off by default.
The arrival also sets up the retirement of the model it supersedes. GitHub will retire MAI-Code-1-Flash on September 10, so teams relying on the older small model should plan to move to 1.1-Flash GitHub retires MAI-Code-1-Flash on September 10.
