Greptile Pricing in 2026 - Plans, Credits, and Real Cost
Greptile pricing verified 2026-10-19 - free Starter, Pro at $30 per seat with 50 credits, how the credit model bites, platform coverage, and what it really costs.
Published:
Last Updated:
Direct answer
Greptile Pro costs $30 per seat per month with 50 credits included and $1 per extra credit (verified 2026-10-19). Starter is free for 1 active developer with 50 credits. The two mechanics that move your bill: a seat is any developer who received a review, and credits are per author and do not pool. Apex-depth reviews cost 10 credits each. Enterprise is quote-only. Verified 2026-10-19.
How Greptile prices its AI code review
Greptile has three published tiers and a credit meter sitting on top of the seat price. The seat price is the easy part. The credit meter and the seat definition are where teams get the forecast wrong, and both changed from how this page described them before (verified 2026-10-19).
Here is the full rate card as published.
| Tier | Price | Credits included | Who it is for |
|---|---|---|---|
| Starter | Free | 50 credits/month | 1 active developer only |
| Pro | $30/seat/month | 50 credits per seat | Teams, unlimited repositories |
| Enterprise | Quote-only | Negotiated | Self-host, BYO-LLM, SSO/SAML |
All three figures verified on Greptile’s published plans 2026-10-19. Extra credits are $1 each on Pro. A 14-day Pro trial is available.
Correction note. An earlier version of this page said Greptile had no permanent free tier, listed only GitHub and GitLab support, described a seat as a developer who opens pull requests, and repeated an up-to-20-percent annual discount. All four were wrong or unverifiable at the 2026-10-19 check. The method we use to re-check claims like these is on our standards page.
What does Greptile actually cost per developer?
Direct answer: Pro is $30 per seat per month including 50 credits, plus $1 per credit over (verified 2026-10-19). A developer at 50 reviews or fewer pays $30. One at 100 standard reviews pays about $80. One using Apex depth at 10 credits per review exhausts the allowance in 5 reviews. Credits do not pool, so team averages mislead.
The credit model is the whole story. Three ways to burn 50 credits:
- 50 standard reviews. One credit each. This is the designed-for case and Greptile has previously said most users stay inside it.
- 5 Apex-depth reviews. Apex costs 10 credits per review (verified 2026-10-19). Turn it on as a default and a seat’s monthly allowance is gone in a week.
- Any mix. There is no separate Apex budget. It draws from the same 50.
Because credits are per author and do not pool, the arithmetic is per person, not per team. Model it that way.
| Developer profile | Reviews/month | Credits used | Monthly cost |
|---|---|---|---|
| Light (1 PR/week) | 4 standard | 4 | $30 |
| Typical (3 PRs/week) | 13 standard | 13 | $30 |
| Heavy (3 PRs/day) | 65 standard | 65 | $45 |
| Very heavy | 120 standard | 120 | $100 |
| Apex-default, moderate | 20 Apex | 200 | $180 |
The last row is the one to notice. Apex depth at 10 credits per review turns a $30 seat into a $180 seat at 20 reviews a month. That is a 6x swing from a single configuration choice, and it is the most important cost lever Greptile exposes. Treat Apex as a deliberate escalation for risky changes, not a default.
What counts as a Greptile seat?
Direct answer: A seat is any developer who received a review, not one who created a pull request (verified 2026-10-19). That inverts the usual mental model. A tech lead who never opens a PR but whose code gets reviewed is billable. Contractors and bots that receive review feedback count. Audit your actual review recipients before signing, not your PR authors.
This is the definition most comparison pages get wrong, including ours until this pass. The practical consequences:
- Your seat count can exceed your PR-author count. Anyone receiving review feedback is in scope.
- Monorepo teams are exposed. If a change touches code owned by several people and they all receive review output, the billable surface grows.
- It is auditable. Pull your last 30 days of review recipients from your Git host and count distinct humans. That is your Greptile seat estimate.
Compare that to how the other common billing units work.
| Tool | Billable unit | Overage | Repository cap |
|---|---|---|---|
| Greptile Pro | Developer who received a review | $1/credit over 50 | Unlimited |
| CodeRabbit Essentials | Developer | $0.25 per reviewed file | 1 linked repo |
| Qodo | Flat team fee to 30 users | $0.012/credit, pooled | Not capped per tier |
| Korbit Pro | Git author | Not published | Not published |
| Macroscope | None, usage only | $0.05 per KB reviewed | Unlimited |
Figures verified 2026-10-19 except Korbit overage, which Korbit does not publish. The CodeRabbit 1-linked-repository cap on Essentials is the single most consequential constraint in this table and almost every comparison omits it. Team covers 5, Advanced 10, Enterprise 20.
Which Git platforms does Greptile support?
Direct answer: GitHub, GitLab, Bitbucket Cloud, Bitbucket Data Center, Gitea, and Perforce via P4 Code Review (verified 2026-10-19). Greptile is the only product in our 51-tool matrix documenting Perforce. Azure DevOps is absent, confirmed at this check. That single gap disqualifies Greptile for most Microsoft-stack shops regardless of price.
Platform coverage is the cheapest disqualifier to check and the one teams skip. For contrast: Gitar covers Azure DevOps Services and Server but not GitHub Enterprise Server (verified 2026-10-18 against Gitar’s own support matrix). Gitar is a paid sponsor of this site and we publish that negative anyway, because the alternative is not worth the sponsorship.
| Platform | Greptile | CodeRabbit | Gitar |
|---|---|---|---|
| GitHub.com | Yes | Yes | Yes |
| GitHub Enterprise Server | Enterprise tier | Yes | No |
| GitLab (SaaS + self-managed) | Yes | Yes | Yes |
| Bitbucket Cloud | Yes | Yes | Yes |
| Bitbucket Data Center | Yes | Yes | Yes |
| Azure DevOps | No | Yes | Yes |
| Gitea | Yes | Not documented | Not documented |
| Perforce (P4) | Yes, uniquely | No | No |
Is Greptile’s 82 percent bug-catch rate real?
Direct answer: The 82 percent is Greptile scoring Greptile on 50 pull requests across 5 repositories, not independently validated. Augment Code’s re-run scored it 45 percent, which is an F-score at 45 percent precision and 45 percent recall, not a catch rate. Augment rebuilt the answer key, so part of the gap is methodology. Augment sells a competing product. Verified 2026-10-19.
We ran this down because the earlier version of this page called the 82 percent independent. It is not, and the correction matters more than the number.
What Augment actually did, in its own words: it expanded and corrected the golden comments by reviewing each pull request manually. So the 82-versus-45 gap is partly a different answer key, not purely self-dealing. Report it that way. It is the more credible version and the more citable one.
The conflicts worth naming:
- Greptile published the 82 percent and sells Greptile.
- Augment published the 45 percent and sold a competing reviewer. Harness acquired Augment’s Cosmos, Auggie CLI and Code Context Engine on 8 October 2026, so that dataset now sits inside Harness.
- The 11 false positives per PR for Greptile versus 2 for CodeRabbit figure circulating widely is Panto’s, from a 17-PR run using an o3-mini classifier. We could not open a primary Panto page carrying that exact pair, so we label it reported, not confirmed.
- The neutral Martian Code Review Bench scored CodeRabbit at 51.2 percent F1 and 49.2 percent precision across roughly 300,000 pull requests, and four vendors have each declared themselves number one on it. Entelligence wins its own benchmark at 47.2 percent aggregate F1.
The line that holds: no party without a commercial stake has published a false-positive rate, precision figure or noise measurement for any shipping commercial AI code reviewer. SWE-PRBench and CR-Bench do publish false-positive data, but they measure bare models in fixed harnesses, not products you can buy. Our method for that distinction is in the noise index methodology.
What does Greptile cost for a 10, 25 or 50 person team?
Direct answer: At Pro list price with all seats inside the credit allowance: $300, $750 and $1,500 per month (verified 2026-10-19). Add $1 per credit over 50 per author. We found no published annual discount at this check. The pre-Series A discount of 50 percent takes a qualifying team to $15 per seat. Seat count tracks review recipients, not PR authors.
| Seats | Monthly at list | Annual at list | Pre-Series A (50% off) |
|---|---|---|---|
| 5 | $150 | $1,800 | $900 |
| 10 | $300 | $3,600 | $1,800 |
| 25 | $750 | $9,000 | $4,500 |
| 50 | $1,500 | $18,000 | $9,000 |
| 100 | $3,000 | $36,000 | $18,000 |
Two honesty notes on this table. First, no annual discount is applied, because we could not verify one at the 2026-10-19 check. If Greptile quotes you a term discount, treat it as negotiated rather than published. Second, overage is excluded because it is per author and therefore unmodellable from a headcount. Pull your own per-developer PR distribution and apply $1 per credit over 50 to each person individually.
Discounts Greptile publishes
| Discount | Who qualifies | Effect |
|---|---|---|
| Pre-Series A | Under $2M revenue, pre-Series A | 50% off |
| Y Combinator | YC companies | One month free plus special pricing |
| Open source | Qualified public projects, apply | Reported free access |
| Annual term | Not published at 2026-10-19 | Unverified |
The startup and YC terms are carried from our 2026-10-18 pass. The open-source program is reported but we did not re-open a primary page for it this round.
Greptile pricing versus the alternatives
Direct answer: Greptile Pro at $30 is above CodeRabbit Essentials at $24 annual and level with Qodo at roughly $30 for up to 30 users (verified 2026-10-19). Greptile wins on repository count and Perforce. CodeRabbit wins on Azure DevOps and public-repo coverage. Qodo wins on pooled credits. Macroscope at $0.05 per KB has no seat fee at all.
| Tool | Entry price | Free option | Credits pool? | Notable constraint |
|---|---|---|---|---|
| Greptile Pro | $30/seat/mo | Starter, 1 dev | No | No Azure DevOps |
| CodeRabbit Essentials | $24/dev/mo annual | Public repos only | n/a, per-file overage | 1 linked repo |
| Qodo | ~$30/mo to 30 users | 14-day trial only | Yes | No rollover on credits |
| Gitar Core | $20/user/mo annual | OSI public repos | n/a | 50-user cap, no GHES |
| Macroscope | $0.05/KB, $0.50 floor | $100 new-workspace credit | n/a | Re-reviews billed each time |
| Korbit Pro | $12/mo annual | Max free for all OSS | n/a | Billed per Git author |
All verified 2026-10-19 unless noted. Two items that invalidate older comparisons: CodeRabbit is no longer free for private repositories (free accounts get 1 summary-only private review per hour), and CodeRabbit Lite was retired on 8 June 2026, so any page quoting a $12 Lite tier is stale. Our CodeRabbit pricing page carries the current ladder.
Versus CodeRabbit. The real decision is not $24 against $30. It is 1 linked repository against unlimited on one side, and Azure DevOps plus public-repo reviews on the other. Full detail in CodeRabbit vs Greptile.
Versus Qodo. Same headline price, opposite credit philosophy. Qodo pools credits across the team at $0.012 each with no rollover; Greptile isolates them per author at $1 each. If your PR volume is lumpy, pooling is worth real money. See Qodo pricing.
Versus Claude Code. Anthropic’s Code Review went GA on 15 October 2026 to Claude Team and Enterprise, which changes the comparison set entirely for teams already paying Anthropic. It is not available on Pro or Max, and it is unavailable to orgs with zero-data-retention, the HIPAA configuration, or CMEK. Detail in Greptile vs Claude Code and Claude code review.
Market context. Gitar was acquired by Sonar in May 2026 and Gitar is a paid sponsor here. Harness has now made 15 acquisitions including DeepSource on 9 September 2026 and Augment’s assets on 8 October 2026. Tracxn counted 351 automated code review startups, 85 funded, and only 37 at Series A or beyond in October 2026. Vendor trajectory is a real input alongside the sticker price. The wider field is ranked in best AI PR review tools.
Greptile funding and company stability
Direct answer: Greptile raised a $25M Series A led by Benchmark in September 2025 at roughly $180M valuation, about $30M total (verified 2026-10-19). Several 2026 articles misdate this to 2026. Greptile is one of only 37 companies in this category at Series A or beyond, out of 351 tracked by Tracxn in October 2026.
The misdating matters because a Series A dated September 2026 implies 13 months more runway than one dated September 2025. If you are underwriting vendor risk on a multi-year contract, use the correct date.
Where Greptile pricing is genuinely unclear
We could not verify these at the 2026-10-19 check, and we would rather name them than guess:
- Annual contract discount. Not published. Earlier coverage claiming up to 20 percent is unverified.
- Greptile Chat at $20 per user per month and the Genius API at $0.45 per request. Both appeared in our earlier pass. Neither was re-confirmed at this check. Treat as unverified.
- Enterprise floor price, seat minimum, or term length. None published.
- Net payment terms, purchase-order process, minimum PO value, support SLA. None published. This is market-wide: zero vendors in our set publish net 30 or net 60 terms, a PO process, or a support SLA. Only two publish any purchasing statement at all, both Enterprise-only.
- Credit consumption for incremental re-reviews. Greptile does not document whether a re-review on a new commit costs a fresh credit. Macroscope does document that re-reviews are billed each time. Greptile does not say either way.
- Per-language precision or recall. No vendor in this market publishes it, and no benchmark breaks it out. Greptile’s claim is all languages supported, which is unfalsifiable.
How we verified this
What we checked, and where. Greptile’s published plan and credit figures were checked against Greptile’s own pricing and documentation on 2026-10-19: the three tiers, the $30 Pro seat, the 50 included credits, the $1 per extra credit, the 10-credit Apex depth, the one-active-developer Starter cap, the 14-day Pro trial, and the platform list including Bitbucket Data Center, Gitea and Perforce with P4 Code Review. The Azure DevOps absence was confirmed at the same check. Competitor prices in the comparison tables were checked against each vendor’s own pricing page on 2026-10-19.
What we corrected on this page. Four claims from the previous version were wrong or unverifiable and have been replaced: that Greptile has no permanent free tier; that it supports GitHub and GitLab only; that a seat means a developer who opens pull requests; and that annual contracts get up to 20 percent off. We also removed the description of the 82 percent benchmark as independent. It is Greptile scoring Greptile on 50 pull requests across 5 repositories.
What we could NOT verify. The annual-contract discount, Greptile Chat at $20 per user per month, the Genius API at $0.45 per request, any Enterprise floor price or seat minimum, the credit cost of an incremental re-review, and any support SLA. Greptile publishes none of these. We also could not open a primary Panto page carrying the 11-versus-2 false-positive pair, so that figure stays labelled reported, not confirmed.
What is carried from a prior pass. The pre-Series A 50 percent discount, the Y Combinator terms, and the open-source program are carried from our 2026-10-18 verification and were not re-opened at source on 2026-10-19. The Augment re-run figures and its answer-key statement are carried from the same pass, sourced to Augment’s published write-up and its evaluation repository.
Conflicts we are disclosing. Gitar appears in two tables on this page and Gitar is a paid sponsor of aicodereview.cc. Its unsupported-GHES status and its lack of any published third-party benchmark are stated anyway. Augment, whose re-run produces the least flattering Greptile number on this page, sold a competing reviewer, and its assets now sit inside Harness. We report who paid for each number. Our scoring and verification method is published at standards.
Recommendations
Criteria stated up front: platform coverage first, because it is a hard disqualifier; then billing shape against your actual PR distribution; then repository count; and only then review quality, because no independent quality measurement exists for any product in this category.
Buy Greptile Pro at $30 if you run Perforce, Gitea or Bitbucket Data Center, you need unlimited repositories on an entry tier, and your PR volume per developer is reasonably even. The Perforce support is genuinely unique in this market and worth a premium on its own if you need it.
Do not buy Greptile if you are on Azure DevOps. It is not supported, and no price makes that work.
Think hard before buying Greptile if one or two engineers open most of your pull requests. Per-author credits that do not pool will overbill exactly that shape of team, and Qodo pooling at $0.012 per credit is the structural answer to it.
Take the Starter tier if you are one developer evaluating on one project. It is free with 50 credits and it is the honest way to test depth before committing a team.
Negotiate Enterprise only if you need self-hosting, bring-your-own-LLM or SAML. Expect no published floor, no SLA and no net terms, and price your own risk accordingly, because that is the market norm rather than a Greptile failing.
Turn Apex off by default. Ten credits a review is a 6x cost multiplier. Reserve it for changes where you would otherwise schedule a human design review.
Further reading
GitarComments are not enough
Gitar applies the fix, validates it in CI, and clears the queue.
See it on your repo Read our independent Gitar reviewFrequently Asked Questions
How much does Greptile cost?
Greptile Pro costs $30 per seat per month and includes 50 credits per seat (verified 2026-10-19). Extra credits are $1 each. A Starter tier is free for 1 active developer with 50 credits per month. Enterprise is quote-only. The number that surprises people is the seat definition - a seat is any developer who received a review, not one who opened a pull request, so your billable count tracks who gets reviewed rather than who submits. Greptile runs a 14-day Pro trial. We found no published annual discount on the pricing page at this check.
Does Greptile have a free tier?
Yes, as of our 2026-10-19 check. Greptile Starter is free for 1 active developer and includes 50 credits per month. That is enough for a solo maintainer or a single engineer evaluating the tool on a side project, and it is a change from earlier coverage, including an earlier version of this page, which said Greptile had no permanent free tier and only a 14-day trial. The trial still exists on Pro. For a team of any size the free tier is not usable, because the one-active-developer cap is the binding constraint, not the credit allowance.
What is a Greptile credit and how many do I need?
One standard review consumes one credit. An Apex depth review consumes 10 credits (verified 2026-10-19). Pro includes 50 credits per seat per month and bills $1 per credit beyond that. The critical mechanic is that credits are per author and do not pool across the team, so a developer opening 100 reviewable changes a month pays roughly $50 in overage while a teammate who opens five wastes 45 credits. Teams with uneven PR distribution should model per-developer volume, not a team average, or the forecast will be wrong in both directions at once.
Do Greptile credits pool across a team?
No. Credits are allocated per author and do not pool (verified 2026-10-19). This is the single most consequential difference between Greptile billing and Qodo billing, where credits are pooled across the team. The practical effect is that Greptile overbills teams with a few heavy committers and underuses the allowance of everyone else. If you have one engineer shipping half the pull requests, their seat will carry nearly all of the overage. Qodo pools, CodeRabbit charges per reviewed file over a cap, and Greptile isolates per author - three genuinely different failure modes.
Which Git platforms does Greptile support?
GitHub, GitLab, Bitbucket Cloud, Bitbucket Data Center, Gitea, and Perforce through P4 Code Review (verified 2026-10-19). Greptile is the only tool in our platform matrix of 51 products that documents Perforce support, which matters for game studios and hardware firms that never left P4. Azure DevOps is absent, confirmed at this check, and that is the gap that disqualifies Greptile for a large share of Microsoft-stack shops. An earlier version of this page listed GitHub and GitLab only, which understated the coverage by four platforms.
Is Greptile's 82 percent benchmark result independent?
No. The 82 percent figure is Greptile scoring Greptile on 50 pull requests across 5 repositories, and it was not independently validated. Augment Code re-ran the evaluation and scored Greptile at 45 percent, which is an F-score at 45 percent precision and 45 percent recall, not a catch rate. Augment also rebuilt the answer key, stating that it expanded and corrected the golden comments by reviewing each pull request manually, so part of the gap is methodology rather than self-dealing. Augment sells a competing product, and Harness acquired its Cosmos, Auggie CLI and Code Context Engine on 8 October 2026.
Is Greptile worth $30 per seat compared with CodeRabbit?
It depends on which constraint binds you. CodeRabbit Essentials is $24 per developer per month billed annually and covers 1 linked repository, which is the constraint most comparisons omit. Greptile Pro is $30 with unlimited repositories but a credit cap that does not pool. On review quality, no party without a commercial stake has published a false-positive rate, precision figure or noise measurement for any shipping commercial AI code reviewer, so neither vendor's quality claim is independently settled. Pick on platform coverage, repository count and billing shape, which are all verifiable.
Does Greptile offer startup or open source discounts?
Greptile publishes a 50 percent discount for pre-Series A companies under $2M revenue, plus one month free and special pricing for Y Combinator companies (carried from our 2026-10-18 pass). That takes an effective Pro seat to roughly $15 per month for a qualifying team, which lands below CodeRabbit Essentials at $24. We could not confirm a published annual-contract discount at the 2026-10-19 check, and earlier coverage on this page claiming up to 20 percent off annual terms should be treated as unverified until Greptile publishes it again.
What does Greptile Enterprise include and what does it cost?
Enterprise is quote-only, which puts Greptile in the majority - 12 of the 15 enterprise tiers we track are quote-only, and only Snyk, Gitar and Entelligence publish an enterprise rate card at all. Enterprise is where self-hosted deployment, bring-your-own-LLM, SSO and SAML live. We could not verify a seat minimum, a floor price, net payment terms or a support SLA, because Greptile publishes none of those. Zero vendors in our set publish net 30 or net 60 terms, a purchase-order process, or a support SLA, so the absence is market-wide rather than specific to Greptile.
Explore More
Tool Reviews
Greptile Review
CodeRabbit Review
Qodo Review
Related Articles
Free Newsletter
Stay ahead with AI dev tools
Weekly insights on AI code review, static analysis, and developer productivity. No spam, unsubscribe anytime.
Join developers getting weekly AI tool insights.
Related Articles
How to Measure AI Code Review False Positives: An Open Protocol
A reproducible protocol for measuring AI code review false positives: taxonomy, PR selection, blind classification and variance. Nobody has run it.
October 18, 2026
guideQodo Merge Pricing: Free vs Pro for PR Reviews in 2026
Qodo Merge pricing verified 2026-10-19 - a flat team fee near $30 a month for up to 30 users, pooled credits at $0.012, and free MIT PR-Agent.
March 4, 2026
guideQodo AI Pricing: Free vs Teams vs Enterprise Plans in 2026
Qodo pricing verified 2026-10-19 - a flat team fee near $30/month for up to 30 users, credits at $0.012 pooled with no rollover, and no free tier.
March 4, 2026