
OpenAI Codex
OpenAI's autonomous coding agent across CLI, IDE, and cloud sandboxes.
About
OpenAI Codex is the company's coding agent, relaunched in May 2025 as a cloud-based system that takes a task, works in an isolated sandbox for minutes to hours, and returns a diff with logs and test citations. It ships in four surfaces: an open-source Rust CLI (Apache-2.0), an IDE extension for VS Code, Cursor, Windsurf and JetBrains, a web dashboard at chatgpt.com/codex, and Codex Cloud for delegating long-running tasks. The CLI also supports MCP tool integrations, subagents for parallel work, web search, and image inputs.
Access is bundled with ChatGPT plans rather than sold standalone: Free and Go cover lightweight CLI use, Plus ($20/mo) unlocks Codex Cloud, and Pro tiers raise limits. You can also point the CLI at an OpenAI API key and pay per token, which drops cloud features. In April 2026 OpenAI moved Codex to a credit system, and limits run on five-hour rolling windows, so heavy sessions can stall mid-afternoon.
Codex competes directly with Claude Code, GitHub Copilot's coding agent, and Devin. Its strength is the tight loop between terminal, IDE, and cloud delegation on one account. Weaknesses: no standalone pricing, opaque credit burn on large repos, and feature churn as models retire (GPT-5.x Codex variants deprecated through 2026). Note the name collision: this is unrelated to the deprecated 2021 Codex model API.
What it can do
- Autonomous multi-step coding in CLI, IDE, and cloud
- Delegated Codex Cloud tasks with diffs and test citations
- Automated pull-request code review
- MCP integrations and subagents for parallel work
- Web search and image inputs in the agent loop
Setup
Install the CLI and sign in with a ChatGPT plan, or point it at an OpenAI API key for pay-per-token CLI use.
Prerequisites
A ChatGPT subscription (Plus or higher for Codex Cloud) or an OpenAI API key.
Use cases
Works with
Trust score — 9/10
Scored 9/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 74; Trending score 94/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.