Claude Code Review reaches GA with per-repo review depth

Code Review is Anthropic's managed reviewer for GitHub pull requests. When it launched in research preview, Anthropic described a team of agents that search a PR for bugs in parallel, verify them to filter out false positives, rank them by severity, and post an overview comment plus inline comments. It does not approve PRs. Reviews were billed on token usage and averaged $15 to $25 at the time, making it the deeper and more expensive option beside the open-source Claude Code GitHub Action.

ClaudeDevs announced that the service is now generally available for Team and Enterprise plans. The concrete new control is review depth per repository: admins choose Quick, Standard (the default) or Deep, so routine codebases can get a lighter pass and critical ones a deeper review. Anthropic says most reviews now finish about 3x faster than during the preview, roughly 20 minutes down to 6 to 10 minutes. It also says reviews find more of the bugs that matter, with findings that state the problem, what triggers it and the fix, but gives no figures for that, and the GA post does not say how the three levels differ in cost.

The launch comes with a time-limited credit offer. An owner of an eligible paid Enterprise organization can claim once between October 15, 2026 and January 15, 2027; the amount, between $10,000 and $50,000, is set at claim time from the organization's active Claude Code seats, and it lasts 30 days from the claim whether or not reviews are running, so Anthropic advises setting up repositories first. Every review counts against it, including reviews of new commits and requested re-reviews. Team plans, organizations billed through a reseller, those connected to GitHub Enterprise Server or a .ghe.com data-residency instance, zero-data-retention organizations and those that claimed an earlier Code Review offer are excluded, and the credits do not cover the self-hosted GitHub Action, the CLI's /code-review command or /ultrareview.