Codex Desktop AppAll GPT models.
Follow the numbered steps in the app. Technical examples are optional.
Before you start
Have your API key and some Outrun credit ready. Keep your key private.
This desktop guide covers Windows and macOS. For Linux, use the Codex CLI guide.
Optional setup helperUse an AI assistant to walk through these steps
Copy the setup instructions into an AI assistant you already use. Review any file changes it suggests before allowing them.
Enter your API key directly in the app or local settings. Keep it out of the assistant’s chat.
Connection settingsShow
Base URL in config.toml: https://api.outrun.to/v1
Key variable: OUTRUN_API_KEY
First model: gpt-6.1-solGet your key from Outrun → API keys and enter it privately in the app or local settings. Test with “Reply with OK only.” Tests use credit; check Activity before retrying.
Setup Steps
- 1Install the Codex app
On Windows, use the Microsoft Store download button above. Install the app and open it.
- 2Sign in with ChatGPT
Open Codex and sign in with your ChatGPT account. A free ChatGPT account is enough; it gives you a Codex helper that can do the setup for you.
- 3Paste the One-Click Setup
Click One-Click Setup - Copy to AI above, paste it into a new Codex chat and send it. Codex edits its own config.toml, points it at Outrun and sets up the OUTRUN_API_KEY variable.
- 4Type your key yourself
When Codex tells you which command or file needs your key, put your key there yourself (it starts with sk-rs-). Never paste the key into the chat.
- 5Restart and test
Fully quit and reopen Codex, start a new chat and send a short prompt. It now runs on Outrun with every GPT model set up (gpt-6.1-sol by default; gpt-6-sol, gpt-6-astra, gpt-5.6-sol and gpt-5.6-terra are ready too) and uses a tiny amount of Outrun Credit.
- 6Manual setup (optional)
Open Commands and settings below only if you prefer setting it up yourself. Keep unrelated settings, use the existing template and enter the key locally. The helper steps above are the primary route.
Manual settings and examplesShow
# Install the Codex desktop app from the Microsoft Store (button above)# Save the key as a user environment variable (replace KEY with your own key).
[Environment]::SetEnvironmentVariable("OUTRUN_API_KEY", "KEY", "User")
# Close and reopen PowerShell / the Codex app so the change takes effect.
# Check in a NEW PowerShell window (prints only the first characters):
$env:OUTRUN_API_KEY.Substring(0, 6)
# If Codex still reports a missing key, use the experimental_bearer_token
# alternative shown in the config template below.Replace KEY with your own key only in your local app, editor or terminal. Never send that key to an AI chat. Copy one command block at a time.
Configuration Template
# Codex config file
# macOS / Linux: ~/.codex/config.toml
# Windows: %USERPROFILE%\.codex\config.toml
# Base URL ends in /v1. Do not append /responses.
model = "gpt-6.1-sol"
model_provider = "outrun"
model_reasoning_effort = "high"
[model_providers.outrun]
name = "Outrun"
base_url = "https://api.outrun.to/v1"
wire_api = "responses"
env_key = "OUTRUN_API_KEY"
# Windows alternative (only if Codex cannot see the OUTRUN_API_KEY variable):
# delete the env_key line above and use this line instead, with your key in place of KEY.
# experimental_bearer_token = "KEY"
# All Outrun GPT models (gpt-6.1-sol is the default above).
# Switch with: codex -p gpt-6-astra (or: codex -m gpt-6-astra)
[profiles."gpt-5.6-sol"]
model = "gpt-5.6-sol"
model_provider = "outrun"
[profiles."gpt-6.1-sol"]
model = "gpt-6.1-sol"
model_provider = "outrun"
[profiles."gpt-6-sol"]
model = "gpt-6-sol"
model_provider = "outrun"
[profiles."gpt-6-astra"]
model = "gpt-6-astra"
model_provider = "outrun"
[profiles."gpt-5.6-terra"]
model = "gpt-5.6-terra"
model_provider = "outrun"Optional: test with a script (uses credit)Show
curl https://api.outrun.to/v1/responses \
-H "Authorization: Bearer $OUTRUN_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"gpt-6.1-sol","input":"Reply with OK only.","max_output_tokens":16}' If something goes wrong
Key missing or rejected (401)
Copy your current key from API keys. Paste it only into the app’s key field or your local key variable, then save and reopen the app.
Not enough available credit (402)
Check Balance and the key’s spending cap. Another request may have temporarily reserved credit; read the message before adding funds.
Wrong URL, model or limit (400/404)
Copy the exact values above. Do not add an operation such as /chat/completions to a Base URL. For GLM, context is 1000000 and maximum output is 65536.
Busy or rate limited (429)
Wait for the active request to finish or the rate limit to reset. Avoid repeatedly submitting the same request.
Interrupted request or uncertain charge
Check Activity first. Keep the request ID for support; never include your key in a message or screenshot.
Model choices and service notes
Easiest setup for beginners: install the Codex desktop app, sign in with a free ChatGPT account, paste the One-Click Setup and Codex configures Outrun for you. All GPT models are set up: gpt-6.1-sol (default), gpt-6-sol, gpt-6-astra, gpt-5.6-sol and gpt-5.6-terra. If a request returns 402, check your account balance and key spending cap. For 429, read the error: wait for the active request to finish or the rate limit to reset. If a paid request is interrupted or its outcome is uncertain, check Activity before trying again; do not automatically repeat it. Choose an available model from Models & Pricing; paused models cannot be used.