Qodo Cover is an open-source AI agent that automatically generates unit tests and raises code coverage. Originally released by CodiumAI and now part of the Qodo platform, it analyses a repository, uses coverage reports to find untested paths, writes meaningful tests in the project’s own style, runs them, and iterates until they pass. It has collected more than 5,600 stars on GitHub.
Writing tests for legacy or fast-moving code is slow, repetitive work, which is why coverage quietly rots. Qodo Cover automates the loop end to end: it scans the codebase, identifies the code that matters and lacks tests, generates candidate tests with an LLM, executes them in your environment, and fixes or discards the ones that fail. It runs from the command line or in CI, works across popular languages and frameworks, and comes from the team behind Qodo’s broader AI code-review platform. Unlike one-shot test generators, Cover validates every test it writes by actually executing it, so the tests that land in your repository are ones that pass and that genuinely exercise the target code. Teams typically run it against the least-covered modules first, review the generated tests like any other pull request, and then let it maintain pressure on coverage as the codebase grows.
Free. Qodo Cover itself is free and open source under the AGPL-3.0 license; Qodo separately sells a broader commercial platform for AI code review and generation.
Engineering teams that want to raise test coverage and catch regressions without spending developer time hand-writing boilerplate tests for existing code.