Starter
Try AI PR review.
- 2 reviewers
- Code Review
- Security Review
- 3 seats included
- 15,000 credits to start, then 5,000/mo
- About 70 reviews to start, then about 25 a month
- Unlimited repos
Codex writes the pull request; IonWarp reviews it on GitHub with models from outside the family that wrote it.
CriticalKnown issue in the Code Review benchmark · caught by DeepSeek V4.1 FlashSee the graded case
Codex code review posts a standard GitHub review when you comment @codex review on a pull request, or on every pull request once automatic reviews are on, and it follows the Code Review Rules in your AGENTS.md files.[1] In GitHub it flags only P0 and P1 issues.[1] Codex comes with every ChatGPT plan, and the reviews it runs on GitHub draw on the plan's Code Review usage.[2] Codex can also run your own code review skill, and OpenAI's repository scanner is Codex Security.
IonWarp adds an independent review. Code Review runs DeepSeek V4.1 Flash, a model from outside the family that wrote the code, and traces each change to the code that calls it; the planner adds Security Review or AI Architecture when the diff needs them. Every finding carries a fix prompt to hand back to Codex; Starter is free and Pro is $49/mo per workspace.
A changed signature, return shape or default that a file the pull request never touched still depends on. In a late-September 2026 read of 386 IonWarp reviews, 52% of the significant findings rested on code the pull request never touched. Keycloak #36880, below, is one of them.[3]
Code ReviewOn Starter, Pro and Max
OpenAI facts are from its public documentation, read on 2026-10-02; numbered notes link to the exact page.
| Criterion | IonWarp | Codex code review |
|---|---|---|
| What it is | A GitHub app that reviews every pull request, adding security, performance and docs reviewers when the diff needs them | Another high-signal review pass on GitHub pull requests that follows your repository guidance[1] |
| When it runs | Every new pull request, and again on each push | When you comment @codex review, or on every pull request with automatic review on[1] |
| What it posts | Findings with a severity, a suggested fix and an AI fix prompt, tracked across pushes until resolved or declined | A standard GitHub code review that flags only P0 and P1 issues[1] |
| Cost | Starter free for 3 seats (15,000 credits on signup); Pro $49/mo per workspace | Included with ChatGPT plans; GitHub reviews draw on the plan's Code Review usage[2] |
| Model | F1 | Precision | Recall | $ per review | Median time |
|---|---|---|---|---|---|
| DeepSeek V4.1 FlashThe model it runs | 49.2% | 49.7% | 48.7% | $0.10 | 6m 56s |
| Muse Spark 1.3A failed review retries on it · best on this benchmark | 56.6% | 65.3% | 50.0% | $0.42 | 3m 18s |
On this benchmark Muse Spark 1.3 scored higher than DeepSeek V4.1 Flash (56.6% F1 against 49.2%). Why IonWarp runs DeepSeek V4.1 Flash: on our own repositories from 1 to 21 September 2026 it filed slightly more P0 and P1 findings per 100 reviews than DeepSeek V4 Pro (14.8 on 549 reviews, against 13.5 on 104) for about a quarter of the cost ($0.034 a review against $0.140).
Measured on the same 50 open-source pull requests, with known issues. See the benchmark
Real IonWarp comments on open-source pull requests, graded against the benchmark's answer key.
Missed byQodo 18Greptile 25Cursor Bugbot 21CodeRabbit 20
4 of 47 · See the benchmark
When V2 is enabled and a client is deleted, the V1-only guard skips setPermissionsEnabled(false) so per-client V2 resources remain and getClientsWithPermission can return stale ids · Code Review
HighKnown issue in the Code Review benchmark · caught by Muse Spark 1.3See the graded case
getSubGroupsCount returns null when cached group is deleted concurrently, violating the Never-null contract of GroupModel · Code Review
CriticalKnown issue in the Code Review benchmark · caught by Muse Spark 1.3See the graded case
OptimizedCursorPaginator crashes with TypeError when using datetime ordering because math.floor(datetime) is invalid · Code Review
HighKnown issue in the Code Review benchmark · caught by Muse Spark 1.3See the graded case
parseRefreshTokenResponse schema uses computed Zod keys that become literal field names no real payload contains, so every shared refresh throws Invalid refreshed tokens - should use z.record or passthrough object · Code Review
HighKnown issue in the Code Review benchmark · caught by Muse Spark 1.3See the graded case
Muse Spark 1.3 is its best model on this benchmark; Code Review runs DeepSeek V4.1 Flash unless you pick another.
One workspace price with seats included. Credits pool across your team. Unlimited repos on every plan.
Try AI PR review.
Get AI PR review right from day 1.
Scale AI PR review.
Custom reviewers + credits · GitHub Enterprise · SSO/SAML + audit logs · Custom security review + SLA
Comment @codex review on the pull request, or turn on automatic review for the repository in Codex settings. IonWarp needs no comment: once its GitHub app is installed, it reviews every new pull request.
Codex code review is one review pass, and in GitHub it reports only P0 and P1 issues. IonWarp runs Code Review on every pull request, adds reviewers for security, performance, docs and architecture when the diff needs them, and tracks every finding across pushes until it is resolved or declined.
Yes. Both post on the same pull request. Keep Codex reviews on and install the IonWarp GitHub app on the repositories you want a second review on.
Yes. Every finding includes a suggested fix and an AI fix prompt. Paste it into Codex or point Codex at the pull request; when the fix is pushed, IonWarp re-reviews and marks each finding resolved, declined or still open.
getSubGroupsCount() returns null when the cached group was deleted in another session and the lazy delegate resolves to null, contradicting GroupModel's documented never-null contract and causing admin REST responses to receive null counts · Code Review