OpenAI has introduced the cloud-based programming AI agent Codex

OpenAI has released a preliminary research version of the cloud agent for software development Codex, capable of performing multiple tasks in parallel.

We’re launching a research preview of Codex: a cloud-based software engineering agent that can work on many tasks in parallel.

Rolling out to pro, Enterprise, and Team users in ChatGPT starting today.

— OpenAI (@OpenAI) May 16, 2025

The assistant operates on the codex-1 platform — an optimized version of ChatGPT o3 for software development.

Initially, the agent is available for users of Pro, Enterprise, and Team plans, with support for Plus and Edu promised by the team “in the near future.”

“Codex can read and edit files, as well as run commands, including test wrappers, linters, and type checkers. Task execution usually takes from one minute to half an hour depending on complexity,” the description states.

Each task is processed in a pre-loaded isolated environment.

Users can check the operation of Codex using quotes, terminal logs, and test results. In case of uncertainty or failures, the agent reports issues for decision-making on further actions, OpenAI clarified.

“Like human developers, agents work best when provided with tailored environments, reliable testing setups, and clear documentation,” emphasized the team.

Let us remind you that in April, OpenAI introduced the local programming assistant Codex CLI.