Anthropic Claude Code Review exits research preview for Team and Enterprise plans on Oct 15
Anthropic Code Review exits research preview and is generally available for Team and Enterprise organizations, with internal data showing merged PRs more than doubled and revert share fell 40% from June to September.
On October 15, Anthropic announced that Claude Code Review exited research preview and is generally available for Claude Team and Enterprise organizations. Code Review runs automated review on new pull requests through GitHub integration, with Quick, Standard, and Deep settings per repository. Anthropic said GA reviews are roughly 3x faster than during preview, dropping from ~20 minutes to 6-10 minutes, but did not publish a service-level objective.
On the same day, Anthropic published 'How the Claude Code team does code review' on claude.dev, detailing how Anthropic internally restructured code review under the condition that Claude writes ~90% of the code. Anthropic reported (without external audit) that merged PRs in the Claude Code repository more than doubled from June to September while the share requiring reverts fell by more than 40%. The team's stated goal: most PRs in Claude Code repositories merge with at most one human in the loop, who can be the author, with no increase in reverts or incidents.
Code Review runs on Anthropic infrastructure in four stages with read-only repository access. Multiple agents look for bugs in parallel; a separate group votes on whether each finding is displayed to humans. Code Review never approves or blocks PRs. Review runs tag findings as Important, Nit, or Pre-existing, and teams can gate merges from CI scripts based on Important count.
Code Review is billed separately from regular subscriptions via usage credits. Eligible Enterprise organizations can claim at least $10,000 in Code Review credits by January 15, 2027.