Bugbot charges you again every time you push. We don't.
Bugbot bills $1.00 to $1.50 per run, and it re-runs on every push. Optibot is a flat seat: re-reviewing the same PR twenty times costs nothing extra.
The seat went away. The meter replaced it.
This is not an argument that Bugbot reviews badly. It is about what happens to a review budget when the price of a review is attached to how often your team pushes. All three figures below come from Cursor's own published material.
The unit is a run, not a PR
Every run is a separate charge, and a ten-PR week is not ten charges. It is however many times your team pushed that week.
Iterating bills again
Incremental Review narrows each pass; it does not stop the charge. Teams report 7 to 10 runs on one PR. One thread describes 28.
The cost dial is Teams-only
Effort levels trade depth against spend, and only Teams plans get them. No rate card, no ceiling, no forecastable upper bound.
One pull request, billed seven times
The first run is the cheap one. Then you address the comments, push, and Bugbot runs again. Teams on Cursor's own forum report 7 to 10 runs on an iterative PR. Drag along the chart to see what a single branch costs by the time it merges.
What one pull request costs as you keep pushing
Bugbot reviews on every PR update by default, and each run is billed. Hover or tap the chart.
Cursor publishes an average of $1.00 to $1.50 per Bugbot run, and its staff have confirmed on the Cursor forum that the figure is per run, not per pull request. Bugbot runs automatic reviews on every PR update by default, so an iterative branch bills each time you push. Optibot charges a flat seat, so re-reviewing the same PR costs nothing extra.
Now scale it to a month
Set your team size, PR cadence, and how many billed runs a typical PR really takes. Optibot is flat by plan; Bugbot moves with all three. The break-even line tells you which side of the trade you are on.
Optibot
$290
per month · $29/user · flat rate
Cursor Bugbot
$384
per month · usage-based est.
Adjust for your team
Engineers
10Team size
PRs / eng / week
8Avg per engineer
Bugbot / run
$1.20Cursor publishes $1.00 to $1.50
Runs / PR
1Bugbot re-runs on every push
You save with Optibot
$1.1k
per year · $94 / month
Break-even: 7 PRs/eng/week at $1.20/run
320
PRS/MO
320
RUNS/MO
1.3×
CHEAPER
Heads up: Cursor bills per run, not per PR, and Bugbot reviews on every PR update by default. An iterative branch bills several times over, so raise Runs / PR to match how your team actually pushes. Cursor has not published a spend cap.
How we calculate: Team size × PRs per engineer per week × 4 working weeks × billed runs per PR × cost per Bugbot run. Bugbot defaults to $1.20 per run (inside Cursor's published $1.00 to $1.50 average) at one billed run per PR; high-effort reviews and extra pushes cost more. Optibot is $29 per user per month , billed annually. Flat pricing that doesn't spike as you ship more, the lowest total cost for actively-shipping teams.
Optibot vs Cursor Bugbot, line by line
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
Optibot
Cursor Bugbot
What teams are saying, on Cursor's own forum
Highlights from public threads on forum.cursor.com. These are posters' opinions, not ours, and some of them are answered by Cursor staff in the same thread. Click through to read the original.
- Cursor Forum: A poster estimates 20 PRs a month at roughly 7 runs each lands at $150 to $200 on top of an existing Pro or Ultra subscription, and calls the move away from the flat $40 seat an unjustifiable hike. Read on Cursor Forum
- Cursor Forum: The same thread notes that with review-on-every-push, a single PR can trigger 7 to 10 Bugbot runs, and that an agent left running could quietly spend hundreds. Read on Cursor Forum
- Cursor Forum: A user asks for an actual rate card, unsure whether the $1.00 to $1.50 figure is per review or per pull request. Cursor staff confirm it is per run and agree the docs should be clearer. Read on Cursor Forum
- Cursor Forum: The same reply confirms effort-level controls are a Teams plan feature, not something rolling out to individual plans. Read on Cursor Forum
- Cursor Forum: Teams report that each review surfaces different bugs, so findings arrive across rounds rather than all at once. One reply describes 28 iterations on a single PR. Read on Cursor Forum
- Cursor Forum: Another poster describes the loop as tedious enough to slow the team down, and raised it in a sprint retrospective. Read on Cursor Forum
- Cursor Forum: A reviewer describes comments being drip fed out over time, forcing repeated passes back over the same pull request. Read on Cursor Forum
- Cursor Forum: The same thread argues the price is hard to justify when a self-hosted review agent gets a broadly similar hit rate. Read on Cursor Forum
Three things that change when reviews stop being metered
Predictable pricing is the headline, but the reason it matters is what your team stops doing once a review no longer costs anything at the margin.
Nobody rations reviews
When runs cost money, someone eventually turns review-on-push off for the busy repo. That is coverage lost where the risk is highest.
Depth without an upgrade
Optibot indexes the whole repo on every push. Cross-file and architectural regressions surface with no effort dial to get wrong.
You can see it working
Cycle time, deployment frequency, AI code ratio and contributor productivity, in the same Pro seat. No second analytics tool.
Where Bugbot is still the right call
No tool wins on every axis. Three cases where Bugbot is the better choice today.
Wider platform coverage
Bugbot reaches Bitbucket and Azure DevOps today. Optibot covers GitHub and GitLab, cloud and self-hosted, with both in development.
One less vendor to onboard
If Cursor is already deployed org-wide, Bugbot is a toggle, not a procurement cycle. Weigh that against a bill that keeps moving.
Low volume costs less
A few tightly scoped PRs a month, merged without much iteration, can land under $29 of runs. See the calculator above.
Push as often as you like
Flat-rate reviews with full codebase context, on GitHub and GitLab. Set up in under 10 minutes, first review on your first PR.
Compare Optibot pricing, or see the CodeRabbit pricing breakdown.
Optibot vs Cursor Bugbot Pricing — Common Questions
How much does Cursor Bugbot cost in 2026?
Bugbot no longer has a seat fee. Cursor retired the $40 per seat per month Bugbot subscription in May 2026, and the change landed for existing customers at their first billing renewal after 8 June 2026. Bugbot is now billed purely on usage: Individual plans draw it from included usage, Teams bills it from on-demand spend. Cursor publishes an average of $1.00 to $1.50 per Bugbot run depending on PR size and complexity, and Cursor staff have confirmed on the Cursor forum that this figure is per run rather than per pull request. Bugbot itself is not sold standalone, so you also need a Cursor plan: Pro from $20/mo, or Teams at $40/user/mo.
Why does a single pull request cost more than one Bugbot run?
Bugbot runs an automatic review on every PR update by default, and each run is billed. Incremental Review, which is on by default, narrows each pass to the changes since Bugbot last looked, but it does not stop that pass from being charged. Teams on the Cursor forum report 7 to 10 runs on a single iterative pull request, and one thread describes 28 review rounds on one branch. That is why the honest unit for budgeting is runs per month, not PRs per month.
Is Optibot cheaper than Cursor Bugbot?
It depends on how much your team pushes. Optibot is flat: Plus is $29/user/mo and Pro is $49/user/mo on annual billing, with no per-review or per-file charge. Bugbot has no seat fee, so a developer who opens a few tightly scoped PRs a month and merges them without much iteration can come in under $29 of runs. Once a developer averages more than about 24 billed runs a month at $1.20 each, the flat seat is the cheaper line item, and unlike usage billing it does not move. Use the calculator on this page with your own team size, PR cadence and runs per PR rather than taking either headline number at face value.
Does Cursor Bugbot support GitLab and Bitbucket?
Yes. Bugbot supports GitHub (including GitHub Enterprise Server), GitLab (including GitLab Self-Hosted), Bitbucket (including Bitbucket Data Center) and Azure DevOps in limited availability. It also does not require the Cursor editor: Bugbot connects to your repositories and reviews pull requests independently of what your engineers code in. Optibot supports GitHub and GitLab, cloud and self-hosted, with Bitbucket and Azure DevOps in development.
Does Bugbot read the whole codebase or just the diff?
Bugbot reads more than the raw diff: Cursor describes it as finding issues in how a change interacts with existing components, not only in the touched lines. But its default review scope is incremental, meaning each run looks at the changes since Bugbot last reviewed, and reviewing the full pull request diff on every push is an opt-in setting. Optibot indexes your entire repository on every push and reasons over that codegraph on every review, which is what catches cross-file and architectural regressions that a change-scoped pass can miss.
Does Bugbot include engineering analytics like DORA or cycle time?
Not in the sense of an engineering productivity platform. Cursor Teams provides Bugbot review analytics: findings per review, billed cost, resolution status, severity, and per-rule acceptance rates. Those measure the reviewer. They do not give you PR cycle time, review turnaround, deployment frequency, AI code ratio or contributor productivity. Optibot Pro includes those in the same seat, so the review tool and the measurement of whether reviews are actually speeding you up are one product rather than two.
How accurate are the figures on this page?
Every hard claim about Bugbot comes from Cursor's own published material: the May 2026 pricing blog post for the retirement of the $40 seat and the $1.00 to $1.50 per-run average, the Bugbot documentation for platform support, review triggers, Incremental Review and effort levels, and the Cursor pricing page for plan prices. The per-run rather than per-PR reading, and the Teams-only scope of effort levels, are both from a Cursor staff reply on the Cursor forum. Reviewer commentary is quoted from public Cursor forum threads and linked to the original. Verify current pricing on Cursor's site before making a decision.
Comparison reflects published pricing as of August 2026. Cursor and Bugbot are trademarks of their respective owner; this page is independent comparative commentary.