Qodo (formerly CodiumAI)

Qodo, the product formerly known as CodiumAI, is an AI-powered code-quality platform built around automated review, test generation, and standards enforcement. It works at three touchpoints: a real-time IDE assistant (VS Code, JetBrains), a pull-request reviewer that posts inline findings on GitHub/GitLab/Bitbucket, and a CLI for terminal-driven workflows. A 'context engine' indexes multi-repo codebases so suggestions reflect actual dependencies and conventions rather than file-local guesses.

The pitch is aimed squarely at engineering teams trying to keep up with AI-generated code: 15+ agentic workflows handle things like bug detection, security checks, test coverage gaps, and rule-based standards enforcement. Qodo claims a 64.3% F1 on Code Review Bench and a #1 Gartner ranking for code understanding, and cites ~800 bugs/month caught at enterprise scale. Pricing is free for individuals and open source, per-seat for Teams, and custom Enterprise with on-prem/air-gapped and single-tenant options.

Under the hood Qodo is model-agnostic, routing across Anthropic Claude, OpenAI, Gemini, and NVIDIA models depending on task. It integrates with Jira, Linear, Monday, and Azure DevOps for ticket context, supports the major language stacks (Python, JS/TS, Java, Go, C#, C++, Ruby, PHP, Swift), and ships with SOC 2 Type II and a zero-data-retention posture for regulated buyers.