$ git diff main -- scorecard.md
19 AI code review tools ranked side by side. 16 publish a rate we can rank. 3 do not and are listed for completeness with no inferred ratesrc. Cheapest published rate is Sourcery at Pro $12/dev/mo (annual)src; most expensive published is Graphite Reviewer at Team $40 per user/month (billed annually, unlimited AI Reviews).
1
| tool | cheapest published | model | platforms | deployments | BYOK |
|---|---|---|---|---|---|
| Sourcery | Pro $12/dev/mo (annual) | Per-developer seat (Pro) or per-user (Team) | github, gitlab | saas | no |
| Bito | Team $12/seat/mo (annual; $15 monthly) | Per-seat on the AI Code Reviews plans, annual or monthly, with 5,000 reviewed lines per seat per month included and $5 per 1,000 lines after | github, gitlab, bitbucket | saas, self-hosted, on-prem | no |
| Git AutoReview | Team $12.49/mo for the whole team (annual; $14.99 monthly) | Per-team subscription, not per seat and not per repo; BYOK on every plan, so model tokens bill to your own provider account | github, gitlab, bitbucket | saas, byok | yes |
| Codacy | Team $18/dev/mo (yearly; $21 monthly) | Per-seat on the Team plan, where a seat is any Git contributor who commits to a private repo; $18/dev/mo yearly or $21/dev/mo monthly | github, gitlab, bitbucket | saas | no |
| GitHub Copilot Code Review | Bundled in Copilot Business $19/user/mo | Bundled in Copilot Business ($19/user/mo) or Enterprise ($39/user/mo); reviews consume AI Credits since Jun 2026 | github, azure-devops | saas | no |
| Claude Code Review | Team standard seat $20/mo annual ($25 monthly), 2-seat minimum | Per-seat Claude plan that includes Claude Code, with Enterprise adding usage at API rates | github | saas, byok | yes |
| Qlty | Pro $20/contributor/mo (annual; $24 monthly) | Per-contributor seat, where a contributor is any user or bot that pushed an analyzed commit, uploaded coverage or logged in within a rolling 30 days | github | saas | no |
| Augment Code Review | Standard $20/mo flat for the team (up to 50 seats, $20 usage included) | One flat monthly price for the whole team, up to 50 seats, with no per-seat charge; usage above the included allowance tops up pay-as-you-go | github | saas | no |
| Macroscope | Team $22/dev/mo (annual, Macroscope v3) | Per-developer seat (Team), repriced June 2026 | github, gitlab | saas, byok | yes |
| CodeRabbit | Pro $24/dev/mo (annual) | Per-developer seat (only devs who open PRs); annual or monthly | github, gitlab, bitbucket, azure-devops | saas, self-hosted | no |
| DeepSource | Team $24 per user/month (billed yearly) | Per-user seat on Team, billed yearly, plus pay-as-you-go AI Review credits metered per 10,000 processed lines of code | github, gitlab, bitbucket, azure-devops | saas, self-hosted, byok | yes |
| Snyk Code | Team $25/contributor/mo (annual, Snyk Code only) | Per-contributing-developer seat (Team) or quote-only (Enterprise) | github, gitlab, bitbucket, azure-devops | saas, self-hosted | no |
| Qodo | Teams $30/user/mo (annual) | Per-user seat with pooled credits ($0.012/credit) | github, gitlab, bitbucket, azure-devops | saas, self-hosted, byok, on-prem | yes |
| Greptile | Pro $30/seat/mo (50 reviews included) | Per-seat with metered overages ($1/extra review) | github, gitlab, bitbucket | saas, self-hosted | no |
| SonarQube | SonarQube Team from $34/month for up to 100k LOC | Per-LOC on both Sonar lines, with Gitar AI Code Review alongside it per user | github, gitlab, bitbucket, azure-devops | saas, self-hosted | no |
| Graphite Reviewer | Team $40 per user/month (billed annually, unlimited AI Reviews) | Per-user seat billed annually; AI Reviews are limited on Hobby and Starter and unlimited from Team up | github | saas | no |
| Cursor BugBot | Usage-based on a Cursor plan, with no published BugBot rate | Usage-based, with no seat fee for BugBot. Individuals draw on the usage included in a paid Cursor plan; Teams bill BugBot from on-demand spend | github, gitlab, bitbucket, azure-devops | saas | no |
| Launchable | Quote-only (contact sales) | Quote-only enterprise contract; no published per-seat list price | github, gitlab, bitbucket, azure-devops | saas, self-hosted | no |
| Diffblue Cover | Quote-only (contact sales) | Quote-only enterprise contract; per-developer or per-codebase | github, gitlab, bitbucket, azure-devops | self-hosted | no |
1The "cheapest published" column is the lowest paid list rate that a real buyer can self-serve today. It is a per-seat rate unless the label says otherwise: Augment Code and Git AutoReview both publish one flat price for the whole team, and Cursor BugBot publishes no rate of its own because it bills from usage. Free open-source tiers are noted on the per-tool pages, not here.
2Platforms reflect first-class support: PR comments, status checks, branch-protection awareness. A vendor that supports a platform via a third-party shim is not listed under that platform here.
3BYOK = the vendor lets you point reviews at your own Claude / Gemini / OpenAI key. A "no" here does not mean the vendor is bad. It means token cost is the vendor's problem, not yours.
4-Accuracy scores (OpenSSF F1, SWE-bench) are not ranked in a single column here: benchmark results bounce per-release and a snapshot rank misleads. See /methodology.
1/calculator - live $/PR for your inputs
2/byok-breakeven - per-tool BYOK crossover
3/methodology - how the columns are sourced