AI-generated: This article was created and published automatically by LinkLoot and was not substantively reviewed by a human editor.

Microsoft brings MAI-Code-1.1-Flash to Copilot at a lower token price

Microsoft's MAI-Code-1.1-Flash is now rolling out in GitHub Copilot. The coding model adds native image understanding, improves performance claims over MAI-Code-1-Flash, and appears in GitHub's Copilot pricing table at $0.20 per million input tokens and $1.20 per million output tokens.

The release is a broad Copilot workflow update, not a standalone model page. GitHub says Free and Student users can receive the model through auto model selection, while Pro, Pro+, Max, Business, and Enterprise users can manually select it where supported.

Key takeaways

• MAI-Code-1.1-Flash is a Microsoft-built lightweight coding model now available in GitHub Copilot.
• GitHub says the model adds native vision support for image understanding and improves coding quality, instruction following, tool use, and performance.
• GitHub's pricing docs list it at $0.20 input, $0.02 cached input, and $1.20 output per million tokens.
• Business and Enterprise admins must enable the MAI-Code-1.1-Flash policy before managed users can access it.

What Microsoft released

Microsoft describes MAI-Code-1.1-Flash as a lightweight, agentic coding model built into GitHub Copilot and VS Code. The company's model card lists a sparse mixture-of-experts architecture with 138 billion total parameters, 5 billion active parameters, text and image inputs, text output, and a 256K-token context length.

Microsoft says the new model improves on the June MAI-Code-1-Flash release with 25% greater token efficiency, 25% faster token streaming, a 22% improvement on Terminal-Bench 2.1 in GitHub Copilot CLI, and a 15% improvement on .NET tasks. Those are vendor-reported results, so teams should treat them as a starting point rather than a substitute for their own repo tests.

Where Copilot users get it

GitHub's changelog says MAI-Code-1.1-Flash is selectable in Copilot CLI, the Copilot cloud agent, the GitHub Copilot app, Copilot Chat on GitHub, Visual Studio Code, Visual Studio, GitHub Mobile, JetBrains IDEs, Eclipse, and Xcode.

Availability depends on plan and policy. Free and Student users do not manually select it; Copilot's auto model selection can choose it for them. Paid individual plans can choose it manually. Business and Enterprise tenants need an administrator to enable the policy first, and that policy starts off by default.

That governance detail matters. A model can be listed as available and still be invisible to a managed developer if the organization has not approved the provider, modality, or model policy.

Price and quality tradeoffs

GitHub's pricing page shows MAI-Code-1.1-Flash at less than one-third of MAI-Code-1-Flash's listed token rates. The old model remains listed at $0.75 input, $0.075 cached input, and $4.50 output per million tokens; the new model is $0.20, $0.02, and $1.20.

That price difference is strongest for repetitive coding interactions: issue triage, first-pass refactors, codebase questions, test generation, documentation changes, and screenshot-to-code experiments. It does not remove the need for review. Microsoft's model card says generated code may be inaccurate, incomplete, or incorrect and should be tested before production use.

The best rollout path is a controlled pilot: enable the policy for a small group, compare accepted edits and failed tests against the team's current lightweight model, then inspect whether vision inputs work consistently across the clients developers actually use.

Source check

• Microsoft AI confirms the August 11, 2026 MAI-Code-1.1-Flash release and vendor-reported efficiency claims.
• GitHub Changelog confirms the Copilot rollout, supported surfaces, plan access, and admin policy requirement.
• GitHub Copilot pricing docs list the MAI-Code-1.1-Flash token prices.
• Microsoft's model card provides architecture, modality, context length, release date, and limitations.
• WindowsForum coverage independently highlights the off-by-default enterprise policy and rollout caveats.