Codex CLI custom API Base URL setup for OpenAI-compatible providers

What Codex CLI needs

When you configure Codex CLI with a custom OpenAI-compatible API provider, Codex CLI needs three things:

• Base URL — The root address of the API gateway. The correct value is https://api.rutaapi.com/v1. Always include /v1.
• API key — A secret token that identifies your RutaAPI account. Create one from the dashboard.
• Model — The identifier of the model that handles your requests. Must be enabled in your RutaAPI account.

Depending on your Codex CLI version, the exact config field or environment variable name may vary. Verify the current Codex CLI documentation for your installed version before configuring a custom provider.

RutaAPI values for Codex CLI

Base URL: https://api.rutaapi.com/v1

API key: Create one from the RutaAPI dashboard at app.rutaapi.com. Keys are shown only once at creation — copy and store them securely.

Model: Choose a model enabled in the RutaAPI app pricing page. Run GET /v1/models to see which models are available in your account.

Environment variable example

Depending on your Codex CLI version and configuration method, you may use environment variables or a config file. The example below shows the pattern:

export OPENAI_API_KEY="YOUR_RUTAAPI_KEY"
export OPENAI_BASE_URL="https://api.rutaapi.com/v1"

If your Codex CLI version uses a different variable name, map the same values to the custom OpenAI-compatible Base URL and API key fields.

Test your key before using Codex CLI

Before running any Codex CLI tasks, verify that your API key and Base URL work together:

curl https://api.rutaapi.com/v1/models \
  -H "Authorization: Bearer YOUR_RUTAAPI_KEY"

If the response returns a model list, your Base URL and API key are working. Use one of the returned model IDs in your Codex CLI configuration.

First request example

Once your key is verified, send a chat completions request to confirm end-to-end connectivity:

curl https://api.rutaapi.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_RUTAAPI_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "YOUR_ENABLED_MODEL_ID",
    "messages": [
      {"role": "user", "content": "Say hello from RutaAPI"}
    ]
  }'

Replace YOUR_ENABLED_MODEL_ID with one of the model IDs from the /v1/models response. Do not hardcode a model name you have not verified.

How to pick a model

Model IDs are returned by the GET /v1/models endpoint. Copy one of the model id values from the response and use it in your Codex CLI config.

Model availability is per-account. If a model ID is not in your /v1/models response, it is not active in your account — add credits or check the app pricing page for enabled models.

Common Codex CLI setup mistakes

Problem	Likely cause	Fix
401 / invalid API key	Wrong key, missing Bearer, expired or disabled key	Create a new key in RutaAPI and test /v1/models
Wrong Base URL	Used dashboard URL or homepage URL instead of API endpoint	Use https://api.rutaapi.com/v1
404 / endpoint not found	Missing /v1 in the Base URL	Always include /v1: https://api.rutaapi.com/v1
Model not found	Model ID is not enabled or typed incorrectly	Check app pricing page for enabled model IDs
429 rate limit	Too many concurrent requests or limited route	Slow down requests or choose another enabled route
503 upstream unavailable	Upstream route, gateway, or provider temporarily unavailable	Retry, switch model route, or check dashboard for status
Environment variable not loaded	Shell session not restarted after export	Restart the terminal or re-export the variables
Shell session not restarted	Config changes not picked up by Codex CLI	Restart Codex CLI or reload the config

Cost and usage tracking

RutaAPI uses prepaid credits. Check the app pricing page for model rates and enabled routes. After test calls, check the usage logs and remaining balance in the dashboard.

View model pricing