AI
GitHub Copilot Code Review Now Available in Azure Repos, Charged Per Review
Microsoft has removed gatekeeping from its AI-powered code review tool for Azure Repos, offering teams an alternative to migrating their repositories to GitHub.
InfoQ
Microsoft has made GitHub Copilot code review publicly available for Azure Repos, eliminating the signup requirement that had limited access since June. The feature automatically reviews pull requests with AI-powered comments and integrates cost tracking into Azure's billing system.
Why Azure Repos Gets the Feature Now#
Microsoft has historically encouraged teams to migrate from Azure Repos to GitHub for access to modern AI and agentic tooling. However, many organizations face substantial barriers to switching—including organizational complexity, regulatory demands, and existing customizations. Rather than force migration, Microsoft is bringing GitHub's code review capabilities directly to Azure DevOps infrastructure.
The timing reflects a pragmatic compromise. According to Microsoft's announcement, the company acknowledged that while some customers are actively planning migrations, others lack the readiness or resources to move and must continue using Azure Repos for their daily workflows.
How Copilot Reviews Work in Azure Repos#
The feature analyzes pull requests and posts review comments automatically. It operates only as an advisor: Copilot always leaves a comment-only review and never approves or blocks merges, meaning it cannot satisfy required-reviewer policies. It will not re-review after subsequent commits unless explicitly asked and does not track or respond to reviewer replies.
Automatic reviews can be triggered through branch policies, with custom instructions applied at the organization, project, repository, or specific path level. The August release also introduced onboarding controls and support for Managed DevOps Pools.
Understanding the Costs and Billing#
Each completed review consumes input, output, and cached tokens, with pricing converted to GitHub AI credits at one credit per cent. The complexity and scope of the analysis varies by pull request size, custom instructions, and model selection—all factors that influence token consumption and final cost.
Charges appear in Azure Cost Management under "GitHub Copilot for AzDO," but with a critical caveat: reports arrive 48 hours after a review completes. Project-level tagging now enables teams to filter and group costs by project rather than viewing only organization-wide totals.

