Honest cost-plus.
AI PR reviews for $9/mo + what we pay Anthropic + 20%. No hidden seat margin, no surprise multipliers — we publish the Anthropic invoice each cycle so the markup math is yours to check.
First 5 reviews on any repo, public or private. Enough to evaluate the bot on a real PR before you decide.
25 private reviews a month. Below the unbudgeted-purchase line — a personal card, no procurement.
100 reviews per seat, pooled across the org. Unlocks the multi-pass cross-check & the design critique.
| Also | Price | Includes |
|---|---|---|
| Enterprise | Custom | Unlimited reviews, SSO, a per-customer SLA. Billed direct. |
| Overage (Solo + Team) | Anthropic cost × 1.20 | Past your tier’s included reviews, usage bills at our Anthropic cost plus a flat 20%. No tier jump, no throttling. |
Public repositories run under the same tier mechanics — there is no separate OSS-free track on the hosted App. (Prefer your own runner? The open-source workflow is free; you bring the Anthropic key.)
Why cost-plus?
We charge a flat tier fee, then pass Anthropic’s API cost through with a fixed 20% markup. Each billing cycle we publish the Anthropic invoice, so you can verify the math directly — no hidden seat margin, no surprise multipliers.
Why not a flat $/review? Anthropic’s per-review cost moves with diff size and which naturalist runs (Beetle on Sonnet, Wasp on Opus, Mantis only on disputes). A flat number either overcharges small PRs or undercharges large ones. Cost-plus is honest about that: your invoice is your invoice, not a pricing model’s guess.
Why a 5-review trial, not unlimited free? Five reviews cover one real PR and a few revisions — enough to judge the bot on your own code. Past that, the economics only work if the bot earns the upgrade.
Pick a plan
Plans are chosen at install time. Install the GitHub App, then manage your tier, usage, and billing on app.cludbug.dev. Change tiers or cancel anytime — month to month, no contract.