Kodus

Kodus is an open-source AI-powered code review platform for engineering teams that want automated, context-aware feedback directly in pull requests across Git-based workflows.

Key capabilities include:

• Context-aware PR analysis that evaluates diffs with awareness of the broader codebase, architecture, workflows, coding standards, and business rules to produce targeted feedback on quality, security, performance, and style.
• Custom review rules written in natural language, plus a rules library; it can also sync existing rule files from tools such as Cursor, Copilot, Claude, and Windsurf.
• Model-agnostic BYOK support for using an existing API key with providers such as OpenAI, Claude, Gemini, Llama, or any OpenAI-compatible endpoint, with zero markup on LLM usage.
• Git provider and work-tracking integrations for GitHub, GitLab, Bitbucket, and Azure DevOps, plus business context connections to Jira, Notion, and Linear; it can turn unimplemented suggestions into tracked issues and support technical debt monitoring.

Pricing: Freemium, with a free open-source Community plan and paid Team plans starting at approximately $10 per developer/month (BYOK) or $30 per developer/month (managed mode). A 14-day free trial is available, and a self-hosted open-source option is supported, where users bring their own API keys and pay model providers directly for usage.

Notable for combining a hybrid AST + LLM approach with real-time processing where source code is not stored, while still supporting broad language coverage (30+ languages), unlimited PRs (subject to Git provider APIs), and flexible cloud or self-hosted deployment.