We raised $3M for agent-native runners

Review Runners vs CodeRabbit for AI code review

Review Runners is StarSling's product for AI code review: review runs as a GitHub Actions job on the starsling-review runner, on your own model and provider key.

CodeRabbit reviews pull requests as a hosted service installed through a GitHub App, and Review Runners run review as a job in your own GitHub Actions workflow, calling your own model through your own provider key. Hosted review bots let you customize their reviewer. Review Runners let you run your own.

GitHub integration
How StarSling works

Last updated 2026-09-27

Head to head

Where review runs
Review RunnersIn a GitHub Actions job in your pull request workflow, on the starsling-review runner.
CodeRabbitIn CodeRabbit's hosted service, installed as a GitHub App that comments on each pull request.
Model and provider key
Review RunnersYou bring your own model through your own provider key, and pay the provider directly.
CodeRabbitAn ensemble of models run by CodeRabbit. The Enterprise self-hosted option connects your own model provider.
Reviewer definitions
Review RunnersSkills and instructions in your repository, loaded from the trusted base branch.
CodeRabbitA .coderabbit.yaml at the repository root, Quiet, Chill and Assertive review profiles, and guideline files such as AGENTS.md, CLAUDE.md and .cursorrules, detected automatically.
Changing the reviewer
Review RunnersA change to a skill takes effect once it merges to the base branch, so a pull request cannot rewrite its own reviewer.
CodeRabbitThe configuration in the branch under review is used for that review, and CodeRabbit learns from your usage over time.
GitHub integration
Review RunnersRuns as a job in the pull request workflow you already have.
CodeRabbitA GitHub App with a CodeRabbit check run that reports review status, plus linters and SAST tools run during review.
Billing model
Review Runners$0.004 a minute for the starsling-review runner, plus your model provider's bill for the tokens.
CodeRabbitPer-developer monthly plans, charged for developers who create pull requests, with free reviews on public repositories.
Security model
Review RunnersThe review job reads the repository (contents: read) and holds pull-requests: write to leave its review as comments on the pull request.
CodeRabbitThe GitHub App holds read and write access to checks, code, commit statuses, issues, pull requests and workflows. CodeRabbit states that it never uses customer code for model training.
Availability
Review RunnersEarly access, starting from the Review Runners page.
CodeRabbitGitHub, GitHub Enterprise Server, GitLab, Azure DevOps and Bitbucket.

Customize a hosted reviewer, or run your own

Hosted review bots let you customize their reviewer. Review Runners let you run your own.

CodeRabbit is configured through a .coderabbit.yaml at the repository root. It detects guideline files such as AGENTS.md, CLAUDE.md and .cursorrules and applies them as review criteria, offers Quiet, Chill and Assertive review profiles, supports custom pre-merge checks on its Team plan and above, and learns from your usage.

With Review Runners, the reviewer is yours: its skills and instructions live in your repository and load from the trusted base branch, and it calls the model you choose through your own provider key.

Where the review runs and who holds the key

CodeRabbit runs review in its hosted service and posts on the pull request through its GitHub App. An Enterprise self-hosted deployment runs CodeRabbit inside your own infrastructure and connects your own model provider.

A Review Runner is a GitHub Actions job in the workflow you already have, on the starsling-review runner. The job reads the repository with contents: read, holds pull-requests: write to leave its review as comments, and calls your model through your own provider key. AI code review in GitHub Actions walks through model choice and reviewer context.

What each one bills

CodeRabbit sells per-developer monthly plans and charges for developers who create pull requests, with free reviews on public repositories.

Review Runners bill $0.004 a minute for the starsling-review runner that executes the review job, and your model provider bills your own key for the tokens. How to run automated code review in GitHub Actions covers the setup and the cost of a review job.

Best fit

Choose Review Runners if

Teams that want AI review to run inside their own GitHub Actions workflow, on their own model and provider key, with reviewer skills versioned in the repository.

Choose CodeRabbit if

Teams that want a hosted review service across GitHub, GitLab, Azure DevOps or Bitbucket, configured through a repository file and review profiles.

FAQ

Is Review Runners a CodeRabbit alternative?

Review Runners is a way to run AI code review yourself: as a GitHub Actions job on the starsling-review runner, calling your own model through your own provider key, with reviewer skills loaded from your repository's trusted base branch. CodeRabbit reviews as a hosted service installed through a GitHub App. Review Runners is in early access, starting from the Review Runners page.

Can CodeRabbit be customized?

Yes. CodeRabbit reads a .coderabbit.yaml from the repository root, detects guideline files such as AGENTS.md, CLAUDE.md and .cursorrules, offers Quiet, Chill and Assertive review profiles, and learns from your usage. Review Runners run reviewers you define in your own repository.

Who pays for the model with Review Runners?

You do, through your own provider key. Your provider bills you for the tokens, and StarSling bills $0.004 a minute for the starsling-review runner that executes the review job.

Sources

Other comparisons

Compare StarSling Runners with the runner and CI options teams evaluate.

References

Get started

Run your own reviewers.

Review Runners run AI code review as a GitHub Actions job, on your model and your provider key, with reviewer skills from your repository. Early access starts on the Review Runners page.