Compare

Excellent vs Greptile: the review bot that really does run code

TREX really does run tests in a sandbox. The difference is the attempt record, not the execution.

Greptile

Greptile in one line

An AI reviewer that builds a graph of your codebase and runs a “swarm” review on each pull request, learning from your team's comments. Its TREX agent writes and runs tests in a sandbox, and it ships a Claude Code plugin, an MCP server and /greploop.

Best for: Teams that want whole-repository review context and a reviewer that can validate a change at runtime without them building the harness first.
Limits: TREX validates a change; it does not keep the history of what the agent tried. There is no attempt record, and no held-out trial when you change the agent's prompt, model or tools.
Stronger than us at: This is the fairest fight on the site and we are not going to pretend otherwise. Greptile is stronger in three places: its code graph gives it whole-codebase review context Excellent does not have; TREX writes the tests it runs, and we write none; and its distribution is at a scale we are nowhere near.
Price 2026-09-23: $30 / seat / mo (Pro, 50 credits per seat)

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 — Greptile on the left, Excellent on the right. Prices are the vendor's own, read on 2026-09-23.

Capability	Greptile	Excellent
Runs code to check a change This is the one review bot where the honest answer to “does it execute anything” is yes.	Yes — TREX writes and runs tests in a sandbox	Yes
Whole-codebase review context Excellent checks a specific attempt against specific checks. It does not build a graph of your repository.	Yes	No
Records each agent attempt and the claim it made	No	Yes
Evidence beyond sandboxed tests	Sandboxed test runs	Tests, builds, browser checks and your existing GitHub Actions or GitLab CI
Tests prompt, model and tool changes on held-out work	No	Yes
Where it runs	Sandbox on Greptile's side; Claude Code plugin and MCP server available	Local CLI and MCP; checks run on your machines
Published price (checked 2026-09-23)	Starter free (50 credits per month); Pro $30 per seat per month with 50 credits per seat, extra credits $1 each; a TREX review costs 3 credits against 1 for a standard review; Enterprise custom	Free to install and run. Consulting starts at $10,000 / month.

Verdict

When to pick which

Greptile is the closest thing in the review category to what we do, and if you want a reviewer that can actually run something, TREX does that today. Pick Excellent when the question stops being “is this diff OK” and becomes “what did the agent try, what proved it, and can someone outside the team replay that later.”

Pick Greptile if

Teams that want whole-repository review context and a reviewer that can validate a change at runtime without them building the harness first.

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.

Still shopping? See the full list of Greptile alternatives →

More matchups

See Excellent against the rest of the category

Excellent vs Tusk

Tusk writes the tests. Excellent proves the attempt — and can take Tusk's tests as the evidence.

Excellent vs CodeRabbit

An LLM reading the diff, versus evidence from checks that actually ran.

Excellent vs Cursor Bugbot

A verifier that works with any agent, versus the agent vendor's own reviewer.

Excellent vs GitHub Copilot code review

The reviewer already in your plan, versus evidence from checks that actually ran.

Excellent vs Qodo

Two products describing agent work in similar words — a rules engine on one side, checks that run on the other.

Excellent vs Braintrust

Braintrust evaluates your LLM app. Excellent verifies the code your agent wrote in your repo.

All comparisons

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.

See how it works