OpenAI Codex Setup

OpenAI Codex CLI is OpenAI’s command-line coding agent. Because KeyMaker is OpenAI-compatible, you can point Codex at KeyMaker and run it against campusGenAI’s models — billed to your sponsor’s funding — instead of a personal OpenAI account.

Prerequisites

• Terminal access (macOS, Linux, or Windows PowerShell)
• A valid KeyMaker API key from your sponsor — see Getting Started

You can check whether Codex is already installed:

codex --version

If you see a version number, skip to Save your API key.

Install Codex CLI

macOS or Linux:

curl -fsSL https://chatgpt.com/codex/install.sh | sh

When prompted Start Codex now?, press y.

Windows (PowerShell):

powershell -ExecutionPolicy ByPass -c "irm https://chatgpt.com/codex/install.ps1 | iex"

Then verify the installation on either platform:

codex --version

If you see a version number, the installation was successful.

Save your API key

Store your KeyMaker key as an environment variable rather than putting it directly in a config file.

Already set up Claude Code?

If you configured Claude Code first, your key is already saved in ~/.claude/settings.json as the ANTHROPIC_AUTH_TOKEN value. You can copy it from there:

cat ~/.claude/settings.json

It’s the same KeyMaker key — only the variable name differs (Codex uses KEYMAKER_API_KEY).

macOS / Linux:

echo 'export KEYMAKER_API_KEY="YOUR_KEY_HERE"' >> ~/.zshrc && source ~/.zshrc

Note

This command is for the zsh shell (the default on macOS). If you use bash, replace ~/.zshrc with ~/.bashrc.

Confirm it was saved:

echo $KEYMAKER_API_KEY

Windows (PowerShell):

[Environment]::SetEnvironmentVariable("KEYMAKER_API_KEY", "YOUR_KEY_HERE", "User")

Restart PowerShell, then verify:

echo $env:KEYMAKER_API_KEY

Treat your key like a password

Anyone with your key can use models billed to your funding source. Never share it or commit it to a repository. See security best practices.

See available models

List the models your KeyMaker account can use:

curl https://thekeymaker.umass.edu/v1/models \
  -H "Authorization: Bearer $KEYMAKER_API_KEY"

Make note of the model name you want to use (for example, gpt5).

Configure Codex

1. Create the Codex configuration directory:

   mkdir -p ~/.codex

2. Open (or create) the configuration file:

   nano ~/.codex/config.toml

3. Add the following configuration, then save. Replace gpt5 with any model from your available list if your team uses a different one:

   model_provider = "keymaker"
   model = "gpt5"
   model_reasoning_effort = "high"
   
   [model_providers.keymaker]
   name = "Keymaker"
   base_url = "https://thekeymaker.umass.edu/v1"
   env_key = "KEYMAKER_API_KEY"
   wire_api = "responses"

   In nano, save with Ctrl + O then Enter, and exit with Ctrl + X. (The same shortcuts work on macOS, Linux, and Windows via WSL/Git Bash.)

4. Confirm the file was saved:

   cat ~/.codex/config.toml

Verify the setup

Start Codex:

codex

Then enter a prompt to confirm it’s routing through KeyMaker:

Say hello and tell me which model provider I am using.

If Codex responds successfully, your configuration is working — all Codex requests now route through KeyMaker, using the models, permissions, and usage controls of the KeyMaker platform.

For existing Codex users

If you’ve previously logged into Codex with a ChatGPT or OpenAI account, log out so Codex uses the KeyMaker provider:

codex logout

If Codex still uses your old configuration after that, remove the config directory and reinstall:

rm -rf ~/.codex

Then reinstall Codex and repeat the setup from Install Codex CLI.

Note

Most new users can skip this section. It’s only needed if Codex keeps using a previous OpenAI configuration after you update config.toml.

Troubleshooting

Authentication errors Confirm your key is available with echo $KEYMAKER_API_KEY, then test it directly:

curl https://thekeymaker.umass.edu/v1/models \
  -H "Authorization: Bearer $KEYMAKER_API_KEY"

If that fails, verify the key is correct, hasn’t expired or been revoked, and re-copy it from the KeyMaker portal if needed.

Model not found List the models available to your account (the curl command above) and update the model value in ~/.codex/config.toml to match.

Codex still uses OpenAI Run codex logout, restart your terminal, and try again. If it persists, rm -rf ~/.codex, reinstall Codex, and repeat the setup.

Configuration file not found Create the directory and file manually: mkdir -p ~/.codex then nano ~/.codex/config.toml.

Note

Available models and provider settings may change over time. If something stops working, re-check the Model Comparison page and your config.toml before reaching out.

Next steps

• Usage and Best Practices — monitoring usage, cost awareness, and security

---

Need help?

• Get help: Contact us
• About campusGenAI: Main site

Your interface may look slightly different depending on your institution’s deployment.