OpenAI announced the launch of Codex, a cloud-based AI software engineering agent designed to handle a wide range of coding tasks directly within ChatGPT.
Codex allows users to delegate software development tasks such as writing features, debugging, running tests, and proposing pull requests.
The feature is now available in research preview for ChatGPT Pro, Team, and Enterprise users, with plans to expand to Plus and Edu subscribers soon.
The tool is powered by codex-1, a model built on OpenAI’s o3 and optimized for software engineering tasks.
According to OpenAI, Codex was trained using reinforcement learning on real-world coding challenges, enabling it to mirror human coding styles, follow instructions precisely, and iteratively run tests until a task passes.
"Codex can perform tasks for you such as writing features, answering questions about your codebase, fixing bugs, and proposing pull requests for review," the company said in a blog post. “Each task runs in its own cloud sandbox environment, preloaded with your repository.”
Users interact with Codex through the ChatGPT sidebar by typing prompts or questions. Codex then operates independently on each task, using tools like linters and type checkers and offering real-time progress updates. Once a task is complete, Codex logs its actions and outputs, allowing users to trace changes, request edits, or directly integrate updates.
The move marks OpenAI’s latest effort to extend the capabilities of its generative AI models beyond chat and into practical software development, as the company seeks to attract enterprise users with productivity-enhancing tools.