Home
Latest reporting
News
Anthropic says Code Review is out of research preview, with faster reviews and tiered settings for Team and Enterprise customers.
Anthropic said its Code Review feature is now generally available for Claude Team and Enterprise organizations, moving the product out of research preview and into broader use for reviewing pull requests before they are merged.
What changed
In a post published on Thu, 15 Oct 2026, Anthropic said the general-availability release focuses on three areas: speed, review quality and flexibility. The company said most reviews now finish about 3x faster than during preview, with typical turnaround reduced from about 20 minutes to 6-10 minutes. It also said the system now produces clearer findings that explain the problem, the trigger and the fix in a few sentences, and that administrators can choose Quick, Standard or Deep review levels per repository.
Anthropic said the tool is designed to help catch bugs before they reach production, noting that teams in preview reported it was finding real issues before a human reviewer looked at a pull request. The company also said it uses Code Review on its own pull requests, which it described as helping identify where reviews slow down internal engineering work.
Access and credits
According to Anthropic, Code Review is available to Claude Team and Enterprise plans. The company said eligible Claude Enterprise organizations can claim at least $10,000 in Code Review credits by January 15, 2027, and that credits expire 30 days after they are claimed. Anthropic said terms apply.
Context
The release lands in a market where AI-assisted code review is becoming a more active product category. Recent GitHub changelog items dated 2026-08-26, 2026-09-11 and 2026-10-08 show ongoing updates to Copilot code review, including public preview support for Azure Repos, analysis updates and new organization billing controls. Background material supplied with the announcement notes that code review tools are intended to supplement human review, and that AI reviewers can miss defects or suggest incorrect fixes, so teams still need to verify findings and test proposed changes.
Anthropic did not provide usage figures, customer counts or performance data beyond its preview-to-general-availability timing claims. It also did not give a date for any next product milestone.
