Alternatives / Greptile
Looking for an open source Greptile alternative? We compared 3 open source AI code review tools by license, supported Git hosts, models, and pricing. They review pull requests with an LLM you choose, and one of them is open core.
Written by Maxime DERAME
Last updated:
| Alternative | Best for | License |
|---|---|---|
| Kodus | Full AI code reviewer for several Git hosts (has an enterprise license file) | AGPL-3.0 + enterprise license |
| Continue | Source-controlled AI checks on pull requests | Apache-2.0 |
| Pullfrog | GitHub Actions bot for review, triage, and fixes | MIT |
We looked for open-source tools that overlap with Greptile's core: reviewing pull requests with an AI that understands context and reports bugs and quality issues. We verified each project's license, recent activity, stars, and supported hosts in its repository. Greptile is a hosted product, so these tools mostly bring your own model and charge for their platform separately. See our methodology for the full criteria.
Kodus and Pullfrog use your own LLM keys, so you pay the model provider directly in addition to any platform subscription. Continue runs checks you define and is not a broad AI reviewer, so it only enforces the rules you write in Markdown.
Kodus is the closest: an AI code reviewer for pull requests that checks quality, security, and performance, with custom rules, support for GitHub, GitLab, Bitbucket, and Azure DevOps, and optional self-hosted runners. Pullfrog is a GitHub Actions bot that runs AI agents for review, issue triage, CI autofix, and merge conflicts, but it is tied to GitHub.
Kodus's repository is AGPL-3.0 and also contains a separate enterprise license file for paid features. Pullfrog is MIT, and Continue is Apache-2.0. Pullfrog's hosted pricing is $0.07 per run plus model costs, with 30 free runs per month.
Continue, Kodus, and Pullfrog all had commits within days of this update.
Greptile is a proprietary hosted service priced per developer, where your code is sent to its infrastructure for indexing and review. Open-source alternatives let you choose your own models, run the reviewer on your own runners, and keep code under your control, at the cost of paying your model provider directly and less turnkey codebase-wide context.
Continue is an open-source coding agent that runs source-controlled AI checks on pull requests. Checks are written in Markdown and stored alongside your code, appear as native GitHub status checks, and surface suggested fixes when code does not meet the standards you defined, enforcing only what you explicitly configure rather than generic review opinions.
Pricing: Free and open source; model provider costs may apply.
Kodus is an open-source AI code reviewer that analyzes pull requests for quality, security, and performance, with custom rules, support for GitHub, GitLab, Bitbucket, and Azure DevOps, and optional self-hosted runners. You bring your own LLM. The repository is AGPL-3.0 and also contains a separate enterprise license file for paid features.
Pricing: Free to self-host; you pay your model provider, and platform and enterprise plans are paid.
Pullfrog is an MIT-licensed, model-agnostic GitHub bot that runs in GitHub Actions and uses AI agents for pull request review, issue triage, CI autofix, and merge conflict resolution, with bring-your-own-key model support. Hosted pricing is $0.07 per run plus model costs, with 30 free runs a month, and open source projects can apply for free access. It is tied to GitHub.
Pricing: 30 free runs a month, then $0.07 per run plus model costs.
| Alternative | Type | License | Self-hosted | Best for |
|---|---|---|---|---|
| Kodus | AI code reviewer | AGPL-3.0 + enterprise license | Yes | Multi-host PR review |
| Continue | PR checks | Apache-2.0 | Yes | Team standards as checks |
| Pullfrog | GitHub bot | MIT | Runs in your Actions | Review, triage, autofix |
Want a full AI reviewer across Git hosts
Kodus reviews pull requests on GitHub, GitLab, Bitbucket, and Azure DevOps with custom rules, using your own LLM.
Want to enforce your own standards as checks
Continue runs source-controlled checks written in Markdown as native GitHub status checks, under Apache-2.0.
Want a GitHub bot for review, triage, and fixes
Pullfrog runs AI agents in GitHub Actions for review, issue triage, CI autofix, and merge conflicts, under MIT.