Codex
OpenAI's coding agent, which works through tasks in its own sandbox and hands back the changes with logs and test results.
By OpenAI · launched 2025-05 · last checked 2026-11-01
Why 4: each task runs in its own container and returns changes with logs and test output; it completes the work alone, but a person must verify and merge it.
What it does. Codex launched in May 2025 as a research preview of a cloud agent: each task runs in an isolated container preloaded with your repository, and the agent returns its changes with terminal logs and test output for you to review. There is also an open-source command-line version. Codex now sits in the ChatGPT desktop app alongside ChatGPT Work, and OpenAI said in June 2026 that it had more than 5 million weekly active users.
Where it falls short. At launch it could not fully test applications with live user interfaces, and a person still has to verify and merge its work. Its usage figures are OpenAI's own and have not been independently checked.
What it can do
Incidents
- None logged.
Closest fiction
Her — Samantha talks to 8,316 people at once. Codex has five million and a task queue.
Related coding agents
- Claude Code Anthropic · autonomy 4/5
- Cursor Anysphere · autonomy 4/5
- Devin Cognition · autonomy 4/5
- Google Antigravity Google · autonomy 4/5
- Grok Build xAI (SpaceXAI) · autonomy 4/5
- Lovable Lovable · autonomy 4/5
History
- 2026-11-01 Added to the index.
Sources
- OpenAI Launches Codex Software Engineering Agent Preview InfoQ, 2025-05-19
- OpenAI touts broadening Codex usage with 5 million weekly active users Constellation Research, 2026-06-02
- Pricing OpenAI
← All agents in the index · Autonomy is rated on what the product does today, on a 1–5 scale. Every fact links to a source.