Last updated 07-23-2026
Category:
Reviews:
Join thousands of AI enthusiasts in the World of AI!
Kodus
Kodus is an open source AI code review platform built around Kody, an agent that reviews pull requests on GitHub, GitLab, Bitbucket, and Azure DevOps. Teams connect their repos, and Kody posts inline comments on diffs covering logic, security, and performance. The product positions itself as a model-agnostic alternative to closed review tools, letting you bring your own LLM API keys from OpenAI, Anthropic, Gemini, and other OpenAI-compatible providers with zero markup on token costs.
Beyond automated reviews, Kodus supports custom Kody Rules so teams can encode their own conventions, plus a deep review mode that runs dedicated bug, security, and performance specialists in parallel. Engineering metrics and a Quality Radar dashboard track review activity across repos. Self-hosted deployment is available via Docker on your own infrastructure, with the core agent released under AGPLv3 on GitHub.
Kodus targets engineering teams that want AI-assisted PR reviews without vendor lock-in on models or opaque bundled AI pricing. A free Community tier covers self-hosted or Kodus-hosted setups with BYOK, while paid Teams and Enterprise plans add hosted infrastructure, unlimited rules, SSO, and compliance features.
Reviews pull requests on GitHub, GitLab, Bitbucket, and Azure DevOps with inline comments on diffs
Model-agnostic: bring your own keys for Claude, GPT-4, Gemini, Llama, or any OpenAI-compatible endpoint
Zero markup on LLM token costs; you pay providers directly at listed API prices
Custom Kody Rules let teams enforce project-specific conventions during reviews
Deep review mode runs dedicated bug, security, and performance specialist agents in parallel
Self-hosted deployment via Docker keeps source code, LLM calls, and audit logs inside your network
14-day free trial on Teams with up to 35 PR reviews and no credit card required
Open source core with self-hosted option keeps code and LLM calls inside your network.
Model-agnostic BYOK means no markup on token costs and full control over which LLM runs reviews.
Custom Kody Rules let teams encode conventions that generic review tools miss.
Teams plan pricing is per developer plus separate LLM API costs, which adds up for large orgs.
Community tier caps Kody Rules at 10 and plugins at 3, limiting customization on the free plan.
Is Kodus free to use?
Yes. The Community plan is free and includes unlimited PRs with your own API key, unlimited users, and up to 10 Kody Rules. Paid Teams starts at $10 per developer per month plus LLM token costs paid directly to your provider.
Which AI models does Kodus support?
Kodus is model agnostic. You can connect Claude, GPT-4, Gemini, Llama, Grok, or any OpenAI-compatible endpoint. You control which model runs each review and pay the provider directly.
Does Kodus store my source code?
No. Kodus processes diffs in real time and does not store your source code on its servers. Self-hosted deployments keep the entire review pipeline inside your own infrastructure.
Which Git providers does Kodus integrate with?
Kodus supports GitHub, GitLab, Bitbucket, and Azure DevOps, including self-managed flavors like GitHub Enterprise Server and GitLab Self-Managed. You can also trigger reviews from the Kodus CLI.
Can I self-host Kodus?
Yes. Kodus offers self-hosted deployment via Docker on your own VM or infrastructure. The core agent is open source under AGPLv3 and available on GitHub with pinned, reproducible images.