Codex
Coding agents · OpenAI Codex, Codex AI, ChatGPT Codex
A coding agent app from OpenAI on GPT models: it reads the project, edits files, runs commands and tests, and prepares pull requests. It runs in ChatGPT, in the desktop app, in the terminal, in the IDE and in the cloud, where tasks run in a separate sandbox; it is one agent everywhere and comes with ChatGPT plans, with limits that depend on the plan. It is built like Claude Code and Cursor. It reads project instructions from AGENTS.md, keeps repeatable processes in skills and connects outside systems through MCP.
Codex vs ChatGPT: Chat and Work
| ChatGPT Chat | ChatGPT Work | Codex | |
|---|---|---|---|
| What it is | a chat | an agent for multi-step tasks | a coding agent |
| What you get | a text reply | a finished deliverable: research, a document, slides, a spreadsheet | code changes, passing tests, a pull request |
| Sees your project | no, only what you paste | files you give it access to | yes, it reads the repo |
| Where to open it | the ChatGPT app and website | the ChatGPT app and website | the ChatGPT desktop app, the terminal (Codex CLI), the IDE, the cloud |
| Model | GPT models | GPT models | the same GPT models |
| Plan | ChatGPT plans | shared limits with Codex | included in ChatGPT plans; limits shared with Work |
What they share and how they differ
| Claude Code | Codex | Cursor | Hermes | |
|---|---|---|---|---|
| Edits code, runs commands | ✓ | ✓ | ✓ | ✓ |
| Rules file | ✓ | ✓ | ✓ | ✓ |
| Own editor | ✗ | ✗ | ✓ | ✗ |
| Terminal | ✓ | ✓ | ✓ | ✓ |
| Tasks in the cloud | ✓ | ✓ | ✓ | ✗ |
| Models from several vendors | ✗ | ✗ | ✓ | ✓ |
| Your own server | ✗ | ✗ | ✗ | ✓ |
✓has✗none
Read more: Claude Code, Codex, Cursor and Hermes Agent: which to pick
Where it appears · 10
- AI agent = LLM + harness · Harness
- MCP server and AI agent for 1C · Harness
- Claude Code, Codex, Cursor and Hermes Agent · Harness
- How to run and analyze interviews about AI use · Research
- ChatGPT Enterprise usage report for a company · Research
More 5
- Harness engineering in practice · Harness
- Approaches to working with a harness · Prompts and review
- A reconcile loop for specs · Specs and context
- AGENTS.md and CLAUDE.md, specs, rules and skills · Specs and context
- Who to interview before buying AI tools · Research