guide

Is CodeRabbit Free? What the 2026 Free Plan Covers

CodeRabbit is free for public repos only as of 2026. Private repos get 1 summary-only review per hour. Full limits, dates, and what changed.

Published:

Last Updated:

Direct answer

CodeRabbit is free for public repositories only as of 2026-10-19. Its pricing page offers free reviews forever for public repos; private repos are not included. A free account gets 1 private PR review per hour, summary-only with no inline comments, plus 3 IDE and 3 CLI reviews per hour. Open-source projects on public repos get paid-tier features at no cost. For private code, the entry price is Essentials at $24/dev/mo annual. Verified 2026-10-19.

The short answer - yes for public repos, effectively no for private ones

CodeRabbit still has a free plan with no expiry and no credit card. What it covers changed in 2026, and the change is the whole story.

CodeRabbit’s pricing page now reads: free reviews forever for public repositories. Private repositories are not part of the free plan. On a private repo a free account gets 1 pull request review per hour, summary-only - a walkthrough of what changed, with no inline comments on the diff - plus 3 IDE reviews and 3 CLI reviews per hour (verified 2026-10-19).

An earlier version of this page said the free tier covered unlimited public and private repositories at 200 files and 4 PR reviews per hour. That is no longer accurate, and we are correcting it visibly rather than quietly. Those numbers described the plan as it stood through 2025 and early 2026. They are still repeated across most of the comparison market, including, until recently, our own best free code review tools roundup.

So the honest answer splits in two:

  • Public repositories: genuinely free, forever, and generous. Open-source projects get paid-tier features including auto-fix, the linter suite and custom instructions at no cost, with the hourly allowance scaling by repository star count.
  • Private repositories: CodeRabbit is a paid tool. Use the 14-day trial to evaluate, then budget $24 per developer per month on Essentials billed annually, or $30 monthly.
CodeRabbit AI code review tool homepage screenshot
CodeRabbit homepage

What exactly does CodeRabbit’s free plan include?

Direct answer: Unlimited reviews on public repositories, PR walkthrough summaries, all four Git platforms, and the IDE extension at 3 reviews per hour (verified 2026-10-19). On private repositories: 1 summary-only review per hour. A single review caps at 150 changed files on both Free and Essentials. Auto-fix, linters, custom instructions and integrations are not on the free plan.

Capability (verified 2026-10-19)FreeEssentials $24/dev/mo annualTeam $48
Public repository reviewsUnlimited, foreverUnlimitedUnlimited
Private repository reviews1 per hour, summary onlyFull inline reviewFull inline review
Linked private repositoriesNot included15
Reviews per hour1 private58
Max files in one review150150300
IDE reviews per hour3HigherHigher
CLI reviews per hour3HigherHigher
PR walkthrough summariesYesYesYes
Auto-fix suggestionsNoYesYes
Built-in linter suiteNoYesYes
Custom instructions in .coderabbit.yamlNoYesYes
Learnable review preferencesNoYesYes
Jira / Linear / SlackNoYesYes
GitHub, GitLab, Azure DevOps, BitbucketYesYesYes
SSO and SAMLNoNoNo - Enterprise only
Audit logsNoNoNo - Enterprise only
API accessNoNoNo - Enterprise only
Self-hostingNoNoNo - Enterprise, 500+ seats

Two rows in that table surprise people. The 150-file cap applies to the free plan and to Essentials equally - paying $24 does not raise it; you need Team at $48 for 300. And SSO, audit logs and API access are Enterprise-only, so a regulated 30-person team cannot buy them at any per-seat price.

AI-powered PR summaries

Every pull request gets an automatically generated walkthrough describing what changed and why, with a file-by-file breakdown. This is the one capability that survives on private repos at the free tier, at 1 per hour. It is genuinely useful for orienting a reviewer on a large diff, and it is not a substitute for inline review.

Inline AI review comments

On public repositories, CodeRabbit’s engine analyzes the diff against repository context and posts inline comments identifying bugs, security issues, style violations and logic errors, with severity labels of Critical, Major, Minor or Trivial (verified 2026-10-19). On private repositories the free plan does not post inline comments at all. That is the single most important difference between the 2025 free plan and the 2026 one.

Support for all four Git platforms

CodeRabbit covers GitHub, GitLab, Azure DevOps and Bitbucket, which is the widest platform coverage in this market (verified 2026-10-19). By comparison, GitHub Copilot is GitHub-only, and Greptile documents GitHub, GitLab, Bitbucket Data Center, Gitea and Perforce but not Azure DevOps. On the free plan that coverage applies, but only to public repositories.

VS Code, Cursor, and Windsurf extension

The IDE extension works on the free plan at 3 reviews per hour (verified 2026-10-19). It reviews staged and unstaged changes before you open a PR. For a solo developer on private code, this plus the CLI’s 3 reviews per hour is the realistic free workflow - 6 local reviews an hour, no inline PR review.

Free plan limits in 2026

Direct answer: 1 private PR review per hour summary-only, 3 IDE reviews per hour, 3 CLI reviews per hour, 150 files maximum per review (verified 2026-10-19). Public repositories are not metered the same way. The old limits of 200 files and 4 PR reviews per hour no longer apply - they described the pre-2026 plan and are the most commonly repeated stale fact about CodeRabbit.

Here is how the hourly ceilings run across the whole ladder, because the free limit only makes sense next to what you are buying:

Tier (verified 2026-10-19)Reviews per hourLinked reposMax files per reviewPrice
Free1 private, summary onlyPublic only150$0
Essentials (was Pro)51150$24/dev/mo annual, $30 monthly
Team (was Pro Plus)85300$48 annual, $60 monthly
Advanced (new)1010300$72 annual
Enterprise1220300Quote only

Three things to note. The hourly limits degrade to 1 per hour under fair-usage enforcement - Essentials drops at 30 and again at 60 reviews per rolling 7 days, so no tier is genuinely unlimited. Work past the hourly limit bills at $0.25 per reviewed file on paid tiers; the free plan has no overage because it queues instead. And the Cloud Coding Agent is billed separately at $0.40 per agent minute.

The ladder itself got more expensive to climb. CodeRabbit retired Lite and Pro Legacy on 8 June 2026 (verified 2026-10-19), which removed the $12 per user per month rung. Combined with the free tier dropping private repos, the step from nothing to a working private-repo review setup went from $12 to $24 per developer per month inside one year.

When is CodeRabbit’s free plan enough?

Direct answer: When your repositories are public. Open-source maintainers get paid-tier features free forever with the allowance scaling by star count, which makes CodeRabbit the most generous OSS offer in this market (verified 2026-10-19). For private code, the free plan is an evaluation surface and not a plan - 1 summary-only review per hour does not support a team.

Open-source projects. Install it and stop reading. Public repositories get free reviews forever, and open-source projects receive paid-tier capability at no cost - auto-fix, linters, custom instructions. CodeRabbit backs this with a $10 million, 12-month commitment it says covers roughly 150,000 open-source projects (its own figure; we did not audit the count). No application process.

Solo developers on public code. Effectively unlimited. Same review engine as paying customers.

Anyone inside the 14-day trial. The trial needs no credit card and gives full paid-tier access (verified 2026-10-19). For private code this is your evaluation window, and unlike 2025 there is no soft landing afterwards.

Teams that only want summaries. If what you actually want is a walkthrough of each PR rather than inline critique, 1 free summary per hour on a private repo may be enough for a low-volume repository. That is a narrow case, and worth naming because it is the only private-repo case the free plan still serves.

When the free plan is not enough

Your code is private and you want inline comments. This is now the whole decision. Free gives you none on private repos.

You run more than one repository. Essentials links 1. A frontend and a backend put you on Team at $48 regardless of headcount - the most consequential and least-reported constraint in CodeRabbit’s pricing.

Your PRs exceed 150 changed files. Free and Essentials both stop there. Team raises it to 300. Our own measurement of 2,896 merged pull requests across 10 repositories found a median of 63 changed lines across 3 files, with p90 at 628 lines - so most teams never hit this, but monorepo refactors do. See how big should a pull request be.

You need SSO, audit logs or API access. Enterprise-only at any seat count.

You need self-hosting. Enterprise, documented at 500 or more seats.

Free alternatives to CodeRabbit for private repositories

Direct answer: Three options are genuinely free on private code (verified 2026-10-19). PR-Agent is MIT-licensed since 23 April 2026 and self-hosts for the cost of your LLM calls. Kodus is AGPLv3, free to self-host at any size, capped at 10 Kody Rules and 3 plugins on Community. SonarQube Cloud’s free tier includes PR analysis up to 50,000 lines on private projects.

PR-Agent was relicensed to MIT on 23 April 2026 - not August, as several write-ups still say - and moved to independent community governance at The-PR-Agent/pr-agent, currently v0.48.0 with Qodo remaining a sponsor. Runs as a CLI, Docker image, GitHub Action or webhook server; covers GitHub, GitLab, Bitbucket, Azure DevOps and Gitea; works with OpenAI, Claude, Gemini or local models via Ollama. The software is free and you pay for tokens, typically $0.02 to $0.10 per review.

Kodus is AGPLv3 as of September 2026 and free to self-host at any team size, including GitHub Enterprise Server, GitLab Self-Managed and any OpenAI-compatible model via vLLM, Ollama, TGI or LiteLLM. Its Community tier caps at 10 Kody Rules and 3 plugins. Its paid Teams tier is $10/dev/mo monthly or $8 annual, Kodus-hosted only, with a 14-day / 35-review trial. One caveat worth stating: Kodus lists SOC 2 compliance as an Enterprise feature and its SOC 2 badge has no documents behind it - we note that without endorsing it.

Codacy runs a Developer tier free forever for individuals, IDE-focused. Team starts at $18/dev/mo yearly ($21 monthly) up to 30 devs, and it is free for open source. Structural catch: Codacy’s AI Reviewer is cloud-and-GitHub-only while GHES requires Codacy Self-hosted, so GHES customers structurally cannot get it.

SonarQube - do not conflate the two products. Community Build is free and self-hosted but analyzes the main branch only and, per Sonar’s own docs, cannot analyze pull requests; Developer Edition is the minimum for branch and PR analysis. SonarQube Cloud’s free tier does include PR analysis, up to 50,000 lines on private projects, and never expires. Community Build also excludes C, C++, Dart, Objective-C, PL/SQL, PowerShell, R, Shell, Swift and T-SQL, plus taint analysis - so it is a non-starter for a C++ shop.

Sourcery is free for open-source repositories; private repos need Pro at $12/seat/mo annual.

GitHub Copilot is the one to be careful about. Copilot Free is the only tier that does not include code review (verified 2026-10-19). Code review starts at Copilot Pro, $10/month. GitHub also retired annual billing entirely on 1 June 2026, so monthly is the only option, and existing annual subscribers drop to Copilot Free at expiry.

For the full list, see best free code review tools, open source AI code review tools and self-hosted AI code review tools.

How CodeRabbit’s free plan compares to Greptile

Greptile is the closest paid comparison, and the honest framing in 2026 is that neither tool is generous on private code.

Greptile’s Starter tier is free for 1 active developer with 50 credits per month (verified 2026-10-19). Above that, Pro is $30 per seat per month with 50 credits included and $1 per extra credit, where a “seat” is any developer who received a review, credits are per author and do not pool, and an Apex-depth review consumes 10 credits rather than 1. Published startup terms: 50% off pre-Series A under $2M revenue, plus one month free and special pricing for YC companies.

What the money buys is full-codebase indexing - the review reasons about cross-file impact rather than the diff alone.

Aspect (verified 2026-10-19)CodeRabbit FreeCodeRabbit Essentials $24Greptile StarterGreptile Pro $30
Private repo inline reviewNoYes, 1 linked repo1 active devYes
Public repo reviewsUnlimited, foreverUnlimitedWithin creditsWithin credits
Billing unit-Per PR creator-Per dev who received a review
Included allowance1 private summary/hr5 reviews/hr50 credits/mo50 credits/mo per author
OverageNone, queues$0.25 per reviewed file-$1 per credit, no pooling
Full codebase indexingNoNoYesYes
Git platformsGitHub, GitLab, Azure DevOps, BitbucketSameGitHub, GitLab, Bitbucket DC, Gitea, PerforceSame
Self-hostingNoNo - Enterprise, 500+ seatsNoEnterprise

On Greptile’s quality claims, be precise. The 82% catch rate is Greptile scoring Greptile on 50 pull requests across 5 repositories, not independently validated. Augment’s re-run on the same repositories scored 45% - and that is an F-score (45% precision / 45% recall), not a catch rate. Augment also rebuilt the answer key, stating it had expanded and corrected the golden comments by reviewing each PR manually, so part of the 82-versus-45 gap is methodology rather than self-dealing. The most neutral number available is from Martian Code Review Bench, published by a company that sells model routing rather than a reviewer: CodeRabbit at 51.2% F1 and 49.2% precision across roughly 300,000 PRs. Four vendors have each declared themselves #1 on that one leaderboard.

The thesis we stand behind, because it survives contact with someone who has read the papers: no party without a commercial stake has published a false-positive rate, precision figure or noise measurement for any shipping commercial AI code reviewer.

For more options, see the CodeRabbit alternatives guide and CodeRabbit vs Greptile.

How we verified this

Checked on 2026-10-19 against CodeRabbit’s published pricing page and documentation plus our own verified fact table. This page needed a larger correction than most, because its central claim had inverted.

What we corrected on this pass:

  1. The free tier no longer covers private repositories. CodeRabbit’s pricing page offers free reviews forever for public repositories and does not include private ones. Free accounts get 1 summary-only private PR review per hour plus 3 IDE and 3 CLI reviews per hour. The previous version of this page stated the free plan covered unlimited public and private repos.
  2. The 200-files-per-hour and 4-PR-reviews-per-hour limits are gone. Those described the pre-2026 plan. Current hourly review limits are 5 / 8 / 10 / 12 across Essentials / Team / Advanced / Enterprise, degrading to 1 per hour under fair usage.
  3. Repository caps added. Essentials links 1 repository, Team 5, Advanced 10, Enterprise 20. The previous version said unlimited repos on every tier.
  4. Tier names. Pro became Essentials and Pro Plus became Team in September 2026 at unchanged prices, and Advanced was added at $72/dev/mo annual. Lite and Pro Legacy were retired on 8 June 2026.
  5. Qodo’s free tier claim removed. The previous version said Qodo Merge offers 75 free PR reviews per month. Qodo has no permanent free tier - a 14-day trial with unlimited reviews, then roughly $30/month flat for up to 30 users plus credits at $0.012 each, pooled with no rollover.
  6. Copilot’s free tier corrected. The previous version said Copilot includes AI code review on its free plan. Copilot Free is the only tier without code review; it starts at Copilot Pro, $10/month.
  7. SonarQube split in two. Community Build cannot analyze pull requests; SonarQube Cloud’s free tier can. The previous version conflated them.
  8. PR-Agent’s licence. MIT as of 23 April 2026, at The-PR-Agent/pr-agent v0.48.0.

What we could NOT verify, named explicitly:

  • The 150,000 open-source projects covered by CodeRabbit’s $10M commitment. That is CodeRabbit’s own count; we did not audit it.
  • Exactly how the open-source hourly allowance scales with star count. CodeRabbit documents that it scales but we could not retrieve the tier boundaries.
  • CodeRabbit’s Enterprise price. Quote-only; no published figure exists. Any dollar amount you see for it is third-party.
  • The Advanced tier’s monthly rate. Only the $72 annual figure is published.
  • Any false-positive rate for CodeRabbit’s free or paid tiers. None exists from a disinterested party. SWE-PRBench and CR-Bench do publish false-positive data, but they measure bare models in fixed harnesses, not products you can buy.
  • Kodus’s SOC 2 status. The badge exists; the documents behind it do not, and Kodus lists SOC 2 compliance as an Enterprise feature. Reported, not endorsed.

Carried from the prior pass (2026-10-18), not re-checked today: the 8 June 2026 Lite and Pro Legacy sunset date; the 500-seat self-hosting threshold; CodeRabbit’s $143M Series C at a $1.5B valuation on 12 August 2026; Greptile’s published startup terms; and the Martian Code Review Bench figures dated 3 March 2026.

Our method for dating claims, handling vendor-funded benchmarks and publishing corrections visibly is on our standards page.

Final recommendation

If your repositories are public, CodeRabbit’s free plan is the best offer in this market and there is no decision to make. Free reviews forever, paid-tier features for open-source projects, no application process, four Git platforms. Install it.

If your repositories are private, treat CodeRabbit as a paid tool and plan accordingly. Run the 14-day trial, which needs no card, and decide inside that window - because the fallback is 1 summary-only review per hour, not a workable free plan. Then price it on two inputs, in this order:

  1. How many active repositories do you have? One means Essentials at $24/dev/mo annual. Two or more means Team at $48. This decides the bill more often than headcount does.
  2. Do your PRs routinely exceed 150 changed files? If yes, you need Team for the 300-file cap regardless of repo count.

If you need free AI review on private code, CodeRabbit is not the answer in 2026. PR-Agent under MIT or Kodus under AGPLv3 are, and the price you pay instead is self-hosting and ops time. That trade is worth making for a small team with DevOps capacity and worth avoiding for a team without it.

Further Reading

Sponsored Why?
Gitar logoGitar

Comments are not enough

Gitar applies the fix, validates it in CI, and clears the queue.

See it on your repo Read our independent Gitar review

Frequently Asked Questions

Is CodeRabbit completely free?

Only on public repositories (verified 2026-10-19). CodeRabbit's pricing page states free reviews forever for public repositories, and private repositories are not included in the free plan. A free account working on a private repo gets 1 pull request review per hour and that review is summary-only, with no inline comments, plus 3 IDE reviews and 3 CLI reviews per hour. There is no time limit and no credit card requirement on the free plan, so it is not a trial. But if your code is private, CodeRabbit is effectively a paid tool and the entry price is Essentials at $24 per developer per month billed annually, or $30 monthly.

Is CodeRabbit free for open source projects?

Yes, and this is the one audience the 2026 change did not hurt (verified 2026-10-19). Public repositories get free reviews forever, and open-source projects get paid-tier features including auto-fix suggestions, the built-in linter suite and custom review instructions at no cost. The hourly review allowance scales with repository star count. CodeRabbit also backs this with a $10 million, 12-month commitment it says keeps the tool free for roughly 150,000 open-source projects, and it has distributed over $600,000 in sponsorships to maintainers. We could not independently audit that project count - it is CodeRabbit's own figure.

What are CodeRabbit's free plan rate limits?

On a private repository, a free account gets 1 pull request review per hour and it is summary-only, plus 3 IDE reviews and 3 CLI reviews per hour (verified 2026-10-19). A single review caps at 150 changed files, the same cap that applies on Essentials. Public repositories are not metered the same way - CodeRabbit's wording is free reviews forever for public repositories. The widely repeated figures of 200 files per hour and 4 pull request reviews per hour described the pre-2026 plan and are now out of date, including in an earlier version of this page.

Did CodeRabbit's free tier change in 2026?

Yes, twice, and both changes raised the cost of starting (verified 2026-10-19). First, the free tier stopped covering private repositories; free accounts now get 1 summary-only private review per hour. Second, CodeRabbit retired the Lite and Pro Legacy plans on 8 June 2026, which removed the $12 per user per month rung on the ladder. The combined effect is that the step from free to paid for a private-repo team went from $12 to $24 per developer per month, and the free tier is no longer a viable permanent landing spot for private code. Most comparison pages still describe the 2025 plan.

What features are missing from CodeRabbit's free plan?

The free plan does not include inline review on private repositories, auto-fix suggestions, the built-in linter suite, custom review instructions in a .coderabbit.yaml file, learnable review preferences, Jira, Linear or Slack integrations, or priority support (verified 2026-10-19). It also does not include SSO, audit logs, API access or self-hosting - but neither do Essentials, Team or Advanced. Those four are Enterprise-only, and self-hosting additionally requires 500 or more seats. On public repositories the free plan is substantially more capable, since open-source projects receive paid-tier features at no cost.

How long does CodeRabbit's free plan last?

Indefinitely. The free plan has no expiration date and no credit card requirement, so it is not a trial (verified 2026-10-19). CodeRabbit separately offers a 14-day trial of the paid features with no card required, and that trial is now the right way to evaluate the tool on private code. The distinction matters more than it used to: when the trial ends, a private-repo team does not fall back onto a usable free plan, it falls back to 1 summary-only review per hour. Plan the purchase decision before the trial expires rather than after.

Does CodeRabbit's free plan work with private repositories?

Barely. A free account gets 1 pull request review per hour on a private repository and that review is summary-only, meaning no inline comments on the diff, plus 3 IDE and 3 CLI reviews per hour (verified 2026-10-19). CodeRabbit's pricing page offers free reviews forever for public repositories and does not include private ones. This is a change from 2025, when the free tier genuinely covered unlimited private repos. Any source telling you CodeRabbit is free for unlimited private repositories at 4 reviews per hour is describing the old plan.

How many repositories can I use on CodeRabbit's free plan?

There is no repository cap on the public side (verified 2026-10-19). On the paid side the caps are strict and widely under-reported: Essentials covers 1 linked repository, Team covers 5, Advanced covers 10 and Enterprise covers 20. That single-repo limit on Essentials is the most consequential constraint in CodeRabbit's pricing, because a team that splits frontend and backend into separate repositories is already past it at $24 and pays $48 on Team. Count your active repositories before you price the seat.

Is CodeRabbit's free plan good enough for a small team?

If your repositories are public, yes - comfortably (verified 2026-10-19). If they are private, no. One summary-only review per hour is not a review workflow for a team of any size, and the useful evaluation path is the 14-day trial followed by Essentials at $24 per developer per month billed annually. At five developers that is $120 per month, and the single linked repository on Essentials will usually bite before the hourly review limit of 5 does. An earlier version of this page recommended running free indefinitely at under five developers; that advice only holds for public code now.

What is the difference between CodeRabbit Free and Essentials?

Free covers public repositories without limit and gives private repositories 1 summary-only review per hour plus 3 IDE and 3 CLI reviews per hour. Essentials at $24 per developer per month billed annually, or $30 monthly, adds full inline review on 1 linked private repository, 5 reviews per hour, auto-fix suggestions, the built-in linter suite, custom review instructions, learnable preferences and Jira, Linear and Slack integrations (all verified 2026-10-19). Both cap a single review at 150 changed files. Note the name: the tier called Pro until September 2026 is now Essentials, at an unchanged price.

Are there free alternatives to CodeRabbit for private repos?

Yes, and they matter more now that CodeRabbit's free tier is public-only (verified 2026-10-19). PR-Agent is MIT-licensed as of 23 April 2026 and self-hosts on your own infrastructure with your own LLM keys, so the software is free and you pay for tokens. Kodus is AGPLv3 and free to self-host at any team size, with a Community tier capped at 10 Kody Rules and 3 plugins. Codacy runs a free-forever Developer tier for individuals. SonarQube Cloud's free tier does include pull request analysis up to 50,000 lines on private projects, unlike SonarQube Community Build, which analyzes the main branch only.

How does CodeRabbit's free plan compare to Greptile?

Neither is generous on private code in 2026 (verified 2026-10-19). CodeRabbit free is public-repo-only with 1 summary-only private review per hour. Greptile's Starter tier is free for 1 active developer with 50 credits per month, then Pro is $30 per seat per month with 50 credits included and $1 per extra credit, where credits are per author and do not pool and an Apex-depth review costs 10 credits. Greptile buys you full-codebase indexing. Be careful with its quality claims: the 82% catch rate is Greptile scoring Greptile on 50 pull requests across 5 repositories, and Augment's re-run scored 45% as an F-score on a rebuilt answer key.

Explore More

Free Newsletter

Stay ahead with AI dev tools

Weekly insights on AI code review, static analysis, and developer productivity. No spam, unsubscribe anytime.

By subscribing you agree to receive the weekly newsletter. Unsubscribe in one click, any time. See our privacy policy.

Join developers getting weekly AI tool insights.

Related Articles