Top News

Anthropic has introduced Code Review, a new tool designed to help developers review code changes more thoroughly. The system uses multiple AI agents to examine pull requests (PRs) and identify potential bugs. Anthropic said the tool is currently available in research preview for Team and Enterprise customers. Announcing the tool, the company shared a blog post writing “Today we're introducing Code Review, which dispatches a team of agents on every PR to catch the bugs that skims miss, built for depth, not speed. It's the system we run on nearly every PR at Anthropic.”


The company said that it has been using the system internally on most of its code changes. According to Anthropic, the tool is intended to support developers by performing deeper reviews and helping reduce the workload on human reviewers while still leaving final approval decisions to engineers.


Anthropic said the new tool was developed after noticing a rise in code output. The company reported that code production per Anthropic engineer has increased by about 200% over the past year.


How Claude Code Review works


When a pull request is opened, Code Review sends multiple AI agents to analyse the code. These agents work in parallel to search for potential bugs.


The agents then verify the detected issues to remove false positives and rank them based on severity. The findings are posted on the pull request as a summary comment along with additional inline comments pointing to specific problems.


The number of agents used depends on the size of the code change. Larger or complex pull requests receive deeper analysis, while smaller changes receive lighter checks. Anthropic said an average review takes around 20 minutes.


Claude Code Review: Cost and control


Code Review optimizes for depth and is more expensive than lighter-weight solutions like the Claude Code GitHub Action. Reviews are billed on token usage and generally average $15–25, scaling with PR size and complexity.


Admins have many ways to control spend and usage:


Monthly organization caps: Define total monthly spend across all reviews

Repository-level control: Enable reviews only on the repositories you choose

Analytics dashboard: Track PRs reviewed, acceptance rate, and total review costs


Claude Code Review: Getting started


Code Review is available now as a research preview in beta for Team and Enterprise plans.


For admins: Enable Code Review in your Claude Code settings, install the GitHub App, and select repositories you’d like to run reviews on.

For developers: Once enabled, reviews run automatically on new PRs. No configuration needed.