Enterprise AI Code Review

Enterprise AI code review you control

Set depth per repo, author or pull request. 55¢ average on Budget, 76¢ average on Balanced.

  • No seat fees
  • Built-in spend controls
  • Set up in under 5 minutes
Pull requestsMode · Review
  • workspaceBalancedApproved
  • repoBudgetMerged
  • labelUltraApproved
  • authorPreciseMerged
  • repoBudgetApproved

The most specific setting wins

Set a workspace default, then override it per repo, author, pull request or review.

Detection Mode settings
This review runsUltraSet by Pull request

Match depth to the risk

Budget for routine changes, Ultra for payments, auth and migrations.

  1. Budget55¢ avgRoutine changes
  2. Balanced76¢ avgThe default
  3. PreciseFewer false positives
  4. UltraPayments, auth, migrations

No unused seats

Your bill follows the code you review, not your headcount.

  • Seats$0
  • Unused licenses$0
  • ReviewsBilled per review

Monthly bill

You're always in control

Set limits so your bill is predictable.

How spend limits work →

Roll it out across your org

Start with a workspace default. Tune repos and teams as you go.

Start free on GitHub

Enterprise AI code review questions

How much does enterprise AI code review cost?

Reviews average 55¢ on Budget and 76¢ on Balanced. Each review is billed by the size of its diff: $0.025 per KB on Budget, $0.05 per KB on Balanced, $0.06 per KB on Precise, $0.20 per KB on Ultra, with a 10 KB minimum per review. There are no seat fees, and new workspaces start with $100 free.

Can teams use different review depth?

Yes. Set a workspace default, then override it for a repository in Settings → Repos, for an author in Settings → Personal, for one pull request with a macroscope-<mode> label, or for one review by commenting @macroscope-app. The most specific setting wins.

How do we keep spend predictable?

Set a monthly spend limit, cap each review and cap each pull request. You can also turn off automatic review for a repository and exclude files from review.

Do we need a config file?

No. Every setting is in the dashboard, a GitHub label or a comment. There is nothing to commit to your repositories.