Anthropic makes Claude Code Review generally available

Anthropic has moved its automated pull request reviewer out of research preview and made it generally available to Claude Team and Enterprise customers. Administrators must enable Code Review in their Claude Code settings, install the GitHub App, and select repositories they'd like to run reviews on. Once configured, Code Review inspects new pull requests and posts findings in GitHub without per-developer setup, according to the GA announcement.

Anthropic is targeting a growing constraint in AI-assisted development: coding agents can produce pull requests faster than human reviewers can assess them. Code Review extends Claude Code from generation into the review queue, where defects, security issues, and maintainability problems still require scrutiny.

Reviews get faster and adjustable

Anthropic reports that most reviews now finish about three times faster than they did during preview. Typical review time has fallen from roughly 20 minutes to between 6 and 10 minutes, shortening the feedback cycle after a pull request opens.

Repository administrators can choose among three review-depth settings:

Anthropic says all three modes maintain its quality threshold, with the setting controlling how extensively the reviewer investigates a change. The company has not published defect-recall benchmarks or a detailed evaluation methodology, so teams will need to measure performance against their own code and review standards.

Review comments have also been shortened in response to preview feedback about noise. Each finding is designed to identify the problem, explain the condition that triggers it, and propose a fix within a few sentences.

Credits come with a 30-day clock

Eligible Claude Enterprise organizations can claim at least $10,000 in Code Review credits. The amount is determined when the offer is claimed and depends on the organization’s active Claude Code seats.

• Claim window: October 15, 2026 through January 15, 2027, ending at 23:59 UTC.
• Credit period: 30 days from the claim time, or until the allocation is exhausted.
• Covered usage: Pull request reviews on github.com through the managed GitHub App.
• Spend limits: Eligible reviews can continue while credits remain, even if the organization has reached its spend limit, set that limit to $0, or exhausted prepaid usage credit.
• Excluded usage: Self-hosted review through GitHub Actions, GitHub Enterprise Server, the /code-review and /ultrareview commands, and other Claude Code activity.

Organizations should configure repositories before claiming the offer because the 30-day period starts immediately. Promotional credits are applied first, and Anthropic says it will not charge eligible reviews beyond the remaining allocation during that period.

The offer applies more narrowly than the product’s general availability. Organizations billed through resellers, connected to GitHub Enterprise Server, or using GitHub Enterprise Cloud with data residency are excluded. Trial and canceled subscriptions, Team, Pro, Max, and Free plans also fail to qualify, as do organizations that claimed Anthropic’s earlier free Code Review offer. Eligible Enterprise subscriptions purchased through cloud marketplaces remain included. The promotion terms contain the full conditions.

Anthropic moves into the review queue

Anthropic uses Code Review on its own pull requests and says that internal deployment helps it identify delays in engineers’ review workflows. The product enters a market that already includes review tools from GitHub Copilot, Graphite, CodeRabbit, and Greptile.

Running the reviewer as a managed GitHub App gives Anthropic a persistent place in the development workflow. Every new pull request in an enrolled repository can flow through Claude automatically, producing usage under the organization’s contract without requiring developers to invoke a command in Claude Code. The promotional credits reduce the cost of establishing that workflow before standard billing begins.

A controlled rollout limits surprises

Teams can activate Code Review by opening Claude Code settings, installing the GitHub App, selecting repositories, and assigning a depth level to each one. High-traffic monorepositories merit close usage monitoring because Deep mode performs more analysis and can consume promotional credits faster.

During an evaluation, teams can compare Claude with an existing reviewer on representative pull requests and track concrete outcomes: valid defects found, false positives, duplicated comments, review latency, developer acceptance, and credit consumption. Security-sensitive repositories may also require an internal review of GitHub App permissions and source-code handling before enrollment.

After 30 days or credit exhaustion, eligible reviews return to normal billing under the organization’s spend limits. Administrators can avoid unplanned charges by setting a budget, monitoring usage during the trial, and deciding which repositories should remain enrolled. Anthropic’s setup guide covers installation and repository configuration.