Glossary
What is OpenAI Codex?
OpenAI Codex is OpenAI's coding agent, available as a command-line tool, IDE extension and cloud agent. It reads and edits code, runs commands and tests, and completes tasks in a repository, locally or in cloud sandboxes. The name was earlier used for a 2021 code model behind GitHub Copilot.
Where it runs
| Surface | Use |
|---|---|
| Codex CLI | Terminal agent on your machine |
| IDE extension | Inside editors like VS Code |
| Cloud | Tasks run in sandboxes, often from ChatGPT |
What it does
Like other coding agents, Codex reads a codebase, plans changes, edits files, runs commands and tests, and can open pull requests. Projects can include an AGENTS.md file with instructions the agent follows.
Beyond code
With tools available, Codex can run scripts and drive other tasks in a project. Launch kits and playbooks written as repository instructions work with Codex the same way they work with Claude Code or Cursor.
Name history
OpenAI first used "Codex" in 2021 for a code-generation model that powered the original GitHub Copilot. The current Codex is an agent product launched in 2025.
Example
A developer runs Codex CLI in a repository and asks it to add tests for a module; Codex writes them, runs the suite, fixes a failing case and summarises the changes.
FAQ
Is Codex the same as GitHub Copilot?
No. Copilot is GitHub's product; Codex is OpenAI's coding agent.
Does Codex run locally?
The CLI and IDE extension run locally; cloud tasks run in sandboxes.
What is AGENTS.md?
A file in a repository with instructions for coding agents.