Compare · category
Excellent vs AI code review: an opinion about a diff is not evidence
Nearly every review bot checks by having a model read the diff. Excellent checks by running things.
AI code review
AI code review, in one line
The bots that comment on your pull requests: CodeRabbit, Greptile, Cursor Bugbot, GitHub Copilot code review, Qodo, Graphite, Devin Review and Sourcery.
- Best for
- Every team shipping pull requests. Review bots catch real bugs, they read well, and most are useful within an hour of install.
- Limits
- Nearly all of them check by having a model read the diff. Devin's own documentation says its review “doesn't execute code”. The exceptions are Greptile's TREX, which writes and runs tests in a sandbox, and GitHub Code Quality, which adds CodeQL — and that is a separate feature from the Copilot reviewer.
- Stronger than us at
- Setup time and readability. A review bot is useful on day one and it tells a human what changed. Excellent is silent until your checks are configured, and it has no opinion about style, naming or whether the change was a good idea.
- Price 2026-09-23
- $12–$72 / dev / mo (Published list prices across the category)
Excellent
Excellent in one line
A verification system for agentic tasks — it records what the agent attempted, runs your checks against the claim, and binds the result to evidence you can open.
- Best for
- Teams that want agent work checked before they trust or promote it.
- Limits
- There is no hosted option. You run it on your machines. Verification evidence stays on the machine that produced it — it does not sync to teammates yet. No SSO, and no Mac app has shipped. It does not write tests, and it does not review your diff.
- Stronger than them at
- Keeping the attempt record, running your own checks as the oracle, and trialling prompt, model and tool changes on held-out work from your repository.
- Price 2026-09-23
- Free to install and run (Consulting from $10,000 / month)
Head to head
The capabilities that decide it
Capability by capability — AI code review on the left, Excellent on the right. Prices are the vendor's own, read on 2026-09-23.
| Capability | AI code review | Excellent |
|---|---|---|
Reads the diff and comments on it | ||
Runs your tests and build | Mostly no. Greptile's TREX is the exception — it writes and runs tests in a sandbox. | |
Records each agent attempt and the claim it made | ||
Independent of the agent | Varies — Bugbot is Cursor's, Copilot code review is GitHub's, Devin Review is Devin's | |
Deterministic analysis | Sonar and GitHub Code Quality add static analysis; the review bots themselves are model-based | Your tests, builds, browser checks and CI decide |
Tests prompt, model and tool changes on held-out work | ||
Published price (checked 2026-09-23) | Sourcery Pro $12 and Team $24 per developer per month annual; Graphite Starter $20 and Team $40 per user per month annual; Greptile Pro $30 per seat per month; Qodo Pro Team $30 per month base to 30 users; CodeRabbit Essentials $24, Team $48, Advanced $72 per developer per month annual; GitHub Copilot code review included with paid Copilot; Cursor Bugbot not published. | Free to install and run. Consulting starts at $10,000 / month. |
Together
They are not mutually exclusive
Excellent is not a review bot and does not want to be. It sits under the review: the bot says what it thinks of the diff, Excellent says what happened when the change was run.
Also on this page
The rest of the field, named
Close enough to belong in this comparison, not different enough to need a page of their own.
Devin Review
VisitReview inside the agent platform that wrote the code. Its own documentation says the review is static analysis and “doesn't execute code”. Devin bills in ACUs; we did not re-check its published tiers, so no price is quoted here.
Sourcery
VisitStatic and LLM review on a budget, with custom rules and security checks. Open Source free forever; Pro $12 per developer per month annual ($15 monthly); Team $24 annual ($30 monthly); Enterprise custom (checked 2026-09-23).
Sonar AI Code Assurance
VisitDeterministic static analysis with stricter quality gates for AI-generated code across 30+ languages, plus an MCP server. It is the incumbent “we assure AI code” story, and it is static only — it has no idea what the agent was trying to do. That page publishes no price (checked 2026-09-23); it links out to a separate plans page.
Verdict
When to pick which
Keep the review bot. It is cheap, it is fast, and a second reader is worth having. Just do not confuse a model's opinion of a diff with proof that the change works — those are different claims, and only one of them survives a question from someone outside the team.
Stay with AI code review if
Every team shipping pull requests. Review bots catch real bugs, they read well, and most are useful within an hour of install.
Add Excellent if
You want agent work captured as attempts, checked against evidence you can open, and promoted only when the result is defensible to someone outside the team.
More matchups
See Excellent against the rest of the category
Trust the evidence, not the agent's confidence
Install Excellent, run it beside the agent you already use, and inspect what the checks actually said.