> ## Documentation Index
> Fetch the complete documentation index at: https://radiusbrowser.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Codex CLI

> Use your existing Codex login, configuration, and provider environment.

| Item          | Behavior                                                                              |
| ------------- | ------------------------------------------------------------------------------------- |
| Use for       | ChatGPT/Codex login, Azure, Amazon Bedrock, gateways, and compatible custom providers |
| Configuration | Your normal `CODEX_HOME`, usually `~/.codex`                                          |
| Credentials   | Your native Codex login or provider environment variables                             |
| Models        | Live results from the Codex app-server                                                |

## How it works

1. Radius launches its bundled Codex app-server with your normal Codex home and
   shell environment.
2. Codex reads your login and `config.toml`.
3. Radius requests the available models.
4. Codex uses the configured model when available. Your Radius selection
   overrides it for the conversation.

If you need to sign in, select **Sign in to Codex**. Radius uses the same Codex
home as the regular CLI, so both can reuse the login.

## Configuration cases

| Situation                                    | What happens                                                                             |
| -------------------------------------------- | ---------------------------------------------------------------------------------------- |
| Codex isn't installed                        | Radius uses its bundled Codex runtime.                                                   |
| `config.toml` doesn't exist                  | Codex uses built-in defaults. Radius can create the file when you open it from Settings. |
| `CODEX_HOME` points elsewhere                | Radius uses `config.toml` from that home.                                                |
| A config file is stored at an unrelated path | Radius doesn't load it automatically. Set `CODEX_HOME` to its directory.                 |
| Your installed Codex is older or newer       | It doesn't affect Radius. Radius uses its bundled version.                               |
| Your config requires a newer Codex version   | Update Radius when a release includes that Codex version.                                |

## Cloud providers

### Google Cloud

Codex doesn't include a dedicated Google Cloud provider. Use Google Cloud only
through a custom model provider whose endpoint supports the APIs that Codex
requires. Configure the endpoint and credential environment variable in your
user-level `config.toml`, then select **CLI** in Radius.

* OpenAI: [Configure custom Codex model providers](https://learn.chatgpt.com/docs/config-file/config-advanced)
* Google Cloud: [Use OpenAI compatibility on Google Cloud](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/start/openai)

### Azure OpenAI

Use **CLI** for Azure. Define the Azure provider, endpoint, API version,
credential environment variable, and deployment in your user-level
`config.toml`. Radius displays the models that Codex reports for that
configuration.

* OpenAI: [Configure custom Codex model providers](https://learn.chatgpt.com/docs/config-file/config-advanced)
* Microsoft: [Use Codex with Azure OpenAI](https://learn.microsoft.com/en-us/azure/foundry/openai/how-to/codex)

### Amazon Bedrock

Codex includes a built-in `amazon-bedrock` model provider. Use **CLI** so Codex
can resolve a Bedrock API key or the standard AWS credential chain from your
native configuration and environment.

* OpenAI: [Use ChatGPT Work and Codex with Amazon Bedrock](https://learn.chatgpt.com/docs/amazon-bedrock)
* AWS: [Get started with OpenAI models and Codex on Amazon Bedrock](https://aws.amazon.com/blogs/machine-learning/get-started-with-openai-gpt-5-6-sol-terra-and-luna-on-amazon-bedrock/)