Skip to content

Codex

OpenAI's coding agent for writing, reviewing and running code across the terminal, IDE and cloud tasks.

Last reviewed Nov 15, 2026Source: https://openai.com/codex/

What it does

Codex is OpenAI's coding agent, available through the Codex app, command-line tool, IDE extension and cloud tasks tied to a repository. It writes code, explains unfamiliar files, reviews changes and runs commands in a controlled environment, with cloud tasks working on a branch you can review before merging. Local work stays in your environment; cloud tasks clone the repository and run in a sandbox. Developers use it for parallel tasks, test writing and code review, and it works alongside ChatGPT rather than replacing it. Access is included in paid ChatGPT plans with limits that depend on the plan tier, and like any agent, its output requires review before it reaches production.

Who it is for

Developers, indie builders, students, and teams on a ChatGPT plan who want an agent writing, reviewing, and testing code across terminal, IDE, and cloud.

What's included

  • Terminal and IDE agent
  • Cloud tasks on repository branches
  • Code review support
  • Sandboxed command execution
  • Available in paid ChatGPT plans

What you can use it for

  • Writing and reviewing code
  • Parallel coding tasks in the cloud
  • Explaining an unfamiliar repository
  • Generating tests

Strengths and limitations

Advantages

  • • Cloud tasks run in parallel
  • • Branches keep changes reviewable
  • • Works from terminal, IDE or web

Limitations

  • • Requires a paid ChatGPT plan
  • • Usage limits vary by tier
  • • Agent output needs review

Pricing

Paid

Included with paid ChatGPT plans, with limits that depend on the plan. Check OpenAI's pricing for current access.

No free plan is listed.

Check official pricing

Try Codex

OpenAI's coding agent for writing, reviewing and running code across the terminal, IDE and cloud tasks.

Visit Codex