Anthropic
Claude Code Review
Claude Code Review is a Anthropic pull-request reviewer on GitHub, GitHub Enterprise Server (self-hosted), billed as usage-based, last verified September 26, 2026. Anthropic's own PR reviewer for teams already on Claude Team or Enterprise — a deep, token-billed review rather than a flat per-seat bot.
Catalog record checked September 26, 2026; vendor pricing and policies can change.
| Specification | Claude Code Review |
|---|---|
| Vendor | Anthropic |
| Git hosts | GitHub, GitHub Enterprise Server (self-hosted) |
| Surface | Multi-agent PR review from Claude Code |
| Billing model | Usage-based |
| Hourly / throughput | Token-billed; admin monthly caps |
| Trial | No separate trial; requires a Claude Team or Enterprise plan. |
| Config / rules | Not verifiedUnverified |
| Inline comment cap | Not verifiedUnverified |
| Self-hosted option | No |
| CI hosts | Not a separate CI product |
| Autofix | Not verifiedUnverified |
| Merge gate | Not verifiedUnverified |
| Free for qualifying OSS | Not verifiedUnverified |
Claude Code Review
Anthropic's own PR reviewer for teams already on Claude Team or Enterprise — a deep, token-billed review rather than a flat per-seat bot.
Best for
- Teams already on Claude Team or Enterprise
- High-stakes PRs where a deeper, costlier review pays off
- GitHub.com or self-hosted GitHub Enterprise Server shops
Watch out
Per-review cost is far above flat-seat reviewers: at an average of $15–25 per review, a busy repo can cost more than a seat-priced bot in days, so set the admin monthly cap before turning on review-after-every-push. GitHub only — *.ghe.com data-residency tenants are not supported, and there is no zero-data-retention option. It is a research preview, so behaviour and billing can change.
How a review works. Multi-agent review of the pull request by Claude. Trigger once after PR creation, after every push, or manually with an “@claude review” comment. GitLab is covered only through a separate CI/CD integration, not this reviewer.
Pricing (verified). Research preview on Claude Team and Enterprise plans. Billed separately through usage credits by tokens — it does not count against the plan's included usage. Anthropic's stated average is $15–25 per review; admins can set monthly spend caps. Not available with zero data retention.
Hourly / throughput. No published hourly PR cap. Each review bills tokens from usage credits (average $15–25 per review per Anthropic), and admins can cap monthly spend. Reviews run once after PR creation, after every push, or on demand, depending on the trigger you configure.
Trial. No separate trial; requires a Claude Team or Enterprise plan.
Watch out. Per-review cost is far above flat-seat reviewers: at an average of $15–25 per review, a busy repo can cost more than a seat-priced bot in days, so set the admin monthly cap before turning on review-after-every-push. GitHub only — *.ghe.com data-residency tenants are not supported, and there is no zero-data-retention option. It is a research preview, so behaviour and billing can change.
Compare Claude Code Review
Common questions
Questions about Claude Code Review
Answered from the verified figures on this page rather than general guidance.