Kimi K2.7 Code reaches general availability in GitHub Copilot July 2026

GitHub enabled Kimi K2.7 Code for Copilot Pro, Pro+, and Max subscribers with gradual rollout. The model runs on Microsoft Azure under usage-based billing at provider list rates. Administrators must toggle the policy for Business and Enterprise tenants before any user can access it. The change supplies the first open-weight option inside the model picker.

No public benchmark numbers accompany the announcement. Prior closed models in Copilot carried documented pass@1 scores on HumanEval and MBPP; Kimi K2.7 Code lacks equivalent published results at launch. Azure hosting removes local inference overhead yet retains the requirement that organizations audit open-weight checkpoints against internal data-governance rules.

The addition expands model choice while shifting cost and compliance decisions to individual users and plan admins. Enterprises that previously restricted Copilot to vetted closed models now face explicit policy gates. Within twelve months, usage telemetry will reveal whether developers select the lower-cost open-weight route for routine tasks or retain proprietary models for higher-stakes code.

Expansion to Copilot CLI, cloud agents, GitHub Mobile, JetBrains, and Xcode follows the initial VS Code release. Further surfaces will activate once quality monitoring confirms stability.