Skip to main content
Update

Claude Code Review Is Now Generally Available, With Reviews About 3x Faster and Review Levels per Repo

Code Review in Claude Code is out of research preview and available to Claude Team and Enterprise organizations. Most reviews now finish in 6-10 minutes instead of about 20, admins can pick Quick, Standard, or Deep for each repo, and eligible Enterprise organizations can claim at least $10,000 in review credits.

By Nattapon YongpaiboonCo-founder, Claude Thailand Community

Anthropic has announced that Code Review in Claude Code is out of research preview and now generally available to organizations on Claude Team and Enterprise plans.

Code Review is a service where Claude reviews the pull requests in your GitHub repositories automatically and leaves comments on the lines where it finds problems.

Announcement image from Anthropic. On the left it reads Code Review is now generally available, with faster reviews that find more of the bugs that matter, and a choice of Quick, Standard, or Deep review levels for each repo, for Claude Team and Enterprise plans. On the right is a diff of src/billing/charge.ts where line 19 changes from const amount = order.total to const amount = order.total * 100. Below the code, a comment from Claude labeled Important says order.total is already in cents, so this charges 100 times the intended amount, and to remove the conversion. A pixel-art character in a hat with a magnifying glass sits at the top right

The image shows an example finding. The code multiplies an amount by 100 even though the value is already in cents. Claude labels it Important and states both the effect, a charge 100 times too large, and the fix, removing the conversion.

What’s new at general availability

  • About 3x faster. Most reviews now finish in 6-10 minutes, down from roughly 20 minutes during the research preview
  • More of the bugs that matter. Findings are concise and state the problem, what triggers it, and the fix
  • Review levels per repo. Admins choose Quick, Standard, or Deep for each repository

Three review levels

Level What it does
Quick A lighter pass. The fastest level, at the lowest cost
Standard Balanced depth, speed, and cost. The default
Deep The most thorough pass. Takes the longest and costs the most

Admins set one level for the organization and can give individual repositories their own, such as Quick for a routine codebase and Deep for a critical one.

How Code Review works

According to the Claude Code documentation, each review runs on Anthropic’s infrastructure in four stages:

  1. Find. Several agents read the diff and the surrounding code in parallel
  2. Verify. Findings are checked against the code and false positives are dropped
  3. Rank. The remaining findings are deduplicated and ranked by severity
  4. Post. Results go on the pull request as inline comments, a summary, and a check run named Claude Code Review

Worth knowing:

  • By default, Code Review focuses on correctness: bugs that would break production, not formatting preferences or missing tests. You can tune what it flags with REVIEW.md and CLAUDE.md files in the repository
  • A review never approves or blocks a pull request. The decision to merge stays with your team
  • Reviews have read-only access to the repository, don’t push commits, and don’t build or run your code
  • Anthropic notes that AI can miss bugs or vulnerabilities, and a review with no findings doesn’t mean the code is free of problems

How to turn it on

Admins turn on Code Review in Claude Code settings, install the Claude GitHub App, and choose the repositories to review.

Developers don’t need to set anything up. Reviews run on new pull requests automatically.

Each repository can be set to review when a pull request opens, after every push, or only on request. Anyone with write access can start a review at any time by commenting @claude review on a pull request.

Who can use it

  • Plans: Claude Team and Enterprise
  • Code hosts: GitHub repositories, including GitHub Enterprise Server and GitHub Enterprise Cloud with data residency
  • Not yet available to organizations with Zero Data Retention, the HIPAA configuration, or customer-managed encryption keys (CMEK)
  • Teams that use Claude Code through Amazon Bedrock, Google Cloud, or Microsoft Foundry still need a Claude organization on a Team or Enterprise plan

Code Review is billed by usage, and more thorough reviews cost more. Admins can set a monthly spend limit.

On other plans you can still have Claude review code, for example with the /code-review command in Claude Code, which works on any plan, or by running Claude in your own CI with GitHub Actions or GitLab CI/CD.

Review credits for Enterprise organizations

Eligible Claude Enterprise organizations can claim Code Review credits:

  • Amount: $10,000 to $50,000, based on the number of active Claude Code seats
  • Claim by: 23:59 UTC on January 15, 2027
  • Valid for 30 days from the claim, or until used up. Unused credits don’t carry over
  • How to claim: an Owner or Primary Owner goes to Organization settings > Claude Code, finds the Code Review section, and clicks Claim. Each organization can claim once
  • Set up your repositories before claiming, because the 30 days start the moment you claim
  • Not covered: reviews on GitHub Enterprise Server, self-hosted Code Review through the GitHub Action, the /code-review and /ultrareview commands in Claude Code, and any usage other than Code Review
  • Not eligible: Team, Pro, Max, and Free plans, organizations on a trial, and organizations that already claimed the earlier free Code Review offer

After the credits expire, Code Review continues under normal billing.

Read the original >

Get it by email

New articles, Claude updates and community event announcements. Sent occasionally, never often enough to annoy you.

The newsletter is written in Thai

Carry on the conversation in our Facebook group

Ask questions, share techniques, show your work and hear about upcoming events. The group is where most of the talking happens.