Claude Code vs Codex.

Two coding agents from two AI companies. What differs, and what does not.

Claude Code is Anthropic's coding agent. Codex is OpenAI's coding agent. Both run in the terminal, and both list other places to run. Claude Code mostly uses Claude models, while Codex uses OpenAI models by default and can connect to other providers. The other big difference is the license. Codex's command line tool is open source, and Claude Code is not. A person still has to review what either one changes.

Facts checked on November 18, 2026.

Words on this page

10 short definitions.
model
The part of an AI tool that uses what it learned to give an answer. Claude and GPT are model families.
provider
A company that offers AI models you can use, such as OpenAI, Anthropic or Google.
local model
An AI model that runs on your own computer instead of on a company's servers.
coding agent
An AI tool that works on code for you. You give it a goal, and it makes the changes.
terminal
A text window where you type commands to your computer. It is also called the command line.
IDE
The program where people write code, also called a code editor. VS Code and the JetBrains editors are well-known ones.
open source
The code is public and its license lets anyone use, change and share it under stated rules.
license
The legal rules for who can use, copy or change the code. MIT and Apache-2.0 are two common open source licenses. Proprietary code can have tighter rules about who can use or change it.
cloud agent
An agent that runs on remote computers you reach over the internet.
branch
A separate line of work in the code. Teams can review changes on a branch before they add them to the main copy.

What is the same

Both are coding agents. You give each a goal in plain words, and each works on your code. Both can also be wrong, so both need a person who reviews the work.

Side by side

The table lines the two tools up row by row. The source links go to the makers' own pages and code.

Claude Code and Codex compared. Checked on 2026-11-18.
FactClaude CodeCodex
What it isA coding agent from Anthropic. It reads your code, edits files and runs commands.A coding agent from OpenAI. The command line version runs on your own computer.
Made byAnthropicOpenAI
Runs inTerminal, VS Code, JetBrains, Desktop app, Web, PhoneTerminal, IDE extension, ChatGPT desktop app, ChatGPT mobile app, ChatGPT on the web, Codex Cloud
Open sourceNo (Proprietary, all rights kept by Anthropic)Yes (Apache-2.0, the command line tool)
ModelsClaude models from Anthropic or through other providers.OpenAI models by default. It can also use other providers, and local models.
First announced or releasedFebruary 2025, as a first previewApril 2025, command line tool announced
Sources

Checked on November 18, 2026. Not published means the maker has not said so where we looked.

What is different, and why it matters

On models, Claude Code mostly uses Claude models from Anthropic. It can reach Claude models through other providers. Codex uses OpenAI models by default. It can also connect to other providers, and to local models.

On the code, Codex's command line tool is open source under the Apache-2.0 license. Anyone can read it and reuse it under that license. Claude Code's license keeps all rights with Anthropic, so it is not open source. This matters if your team needs to read, reuse or change the code.

On where it runs, both work in the terminal. Claude Code also lists VS Code, JetBrains, a desktop app, the web and a phone. Codex also lists an IDE extension, the ChatGPT apps and a cloud version.

How to pick

Begin with the models you already use and trust. Then ask whether anyone needs to read or change the tool's own code. Last, check where you want it to run, and compare that with the lists above.

Still unsure? Try both on one small job that cannot hurt anything, and read the changes before anything ships.

If you would rather have the work done

These tools are for people who write software. Would you rather have the work done than learn a tool? We build and run AI agents for teams, and a person approves what matters.

Questions people ask

Which one should my team pick?

We do not rank them. Compare the rows that matter to you, such as the models, the license and where each one runs. Then try both on something small.

Which one is open source?

Codex's command line tool is open source under the Apache-2.0 license. Claude Code is not. Anthropic keeps all rights to it.

Do they run on my own computer?

The tool itself runs in your terminal. The model it asks usually does not, unless you connect a local model. Codex supports that, and we did not find the same for Claude Code. Read each maker's data terms before you use either one on private work.

How do I check what an agent changed?

Ask the agent to work on a branch, which is a separate line of work in the code. Read the changes before you accept them. Then add them to the main copy only if a trusted person approves.

When were these facts last checked?

On the date shown at the top of the page and under the table. Each fact lists its source. These tools change fast, so check the date before you rely on a detail.

Keep reading