On this page

Get-Codex Documentation

Important: Codex Desktop, Codex CLI, and API calls must use https://api.get-apikey.com/v1 as the API Base URL, including /v1.

Quick Start

1. Install and open VS Code

Install Visual Studio Code first. Codex will use it later to open and edit projects. After installation, open VS Code once and confirm that it starts normally.

Download VS Code

3. Open Codex and update configuration

3.1 Get your API key
Create or copy your key on the API Key page. In the next step, paste it into experimental_bearer_token in config.toml.

Go to API Key page

3.2 Update config.toml
The Settings entry is different on Mac and Windows. Choose your system in the tabs below.

Enter your API key: Replace your-api-key in experimental_bearer_token = "your-api-key" with the actual key copied from the API Key page. Keep the double quotes, then save the configuration.

3.2 Update config.toml

  • In Codex Desktop, go to the lower-left Settings → Config → Open config.toml.
  • Clicking Open config.toml opens VS Code.
  • Update the top-level settings shown below, and replace the entire existing [model_providers.get-codex] section with the one below so that old authentication settings are removed. Keep other providers, sandbox, MCP, project, and desktop settings in the file. In VS Code, press Command+S to save. You can also close the VS Code window and click Save when prompted.

Enter your API key: Replace your-api-key in experimental_bearer_token = "your-api-key" with the actual key copied from the API Key page. Keep the double quotes, then save the configuration.

Open the Settings menu from the lower-left corner of Codex Desktop
Open config.toml from the Codex Desktop Config page
Paste and save config.toml in VS Code

Important: You must use https://api.get-apikey.com/v1 as the API Base URL. Set base_url to this exact address, including /v1.

model = "gpt-5.5"
model_reasoning_effort = "high"
model_provider = "get-codex"

[model_providers.get-codex]
name = "Get-Codex"
base_url = "https://api.get-apikey.com/v1"
wire_api = "responses"
requires_openai_auth = true
experimental_bearer_token = "your-api-key"

If you cannot access Codex Settings yet, open or create the configuration file directly in VS Code at the default location below:

Default path

~/.codex/config.toml

4. Restart Codex and complete the test

After saving config.toml, fully quit and reopen Codex Desktop. Start a new conversation and send this message:
hi
Codex Desktop sends hi and returns a response after saving the configuration
If Codex returns a normal response, you can continue using it.

5. Configure Codex CLI (optional, for command-line interaction)

If you only use Codex Desktop, you can skip this step. If you want to use Codex from a terminal, configure it as follows.

1. Check Node.js / npm
Codex CLI is installed with npm, so first confirm that node and npm are available:

Check whether it is already installed

node -v
npm -v

Install Node.js only if needed

brew install node

Verify again after installation

node -v
npm -v
If the first check already prints valid versions, skip Node.js installation and continue directly to Codex CLI. You can also use the official Node.js download page to install the latest LTS release.
2. Install Codex CLI
After Node.js is ready, install Codex CLI and verify its version:

Install Codex CLI

npm i -g @openai/codex

Verify the version

codex --version
If npm is not available here, go back to the Node.js setup step above first.
3. Configure your API key
Open config.toml at the location below, creating the file if needed. Update the top-level settings shown below, and replace the entire existing [model_providers.get-codex] section with the one below so that old authentication settings are removed. Keep other providers, sandbox, MCP, project, and desktop settings in the file. If you have already configured Desktop above, reuse the same file.

Default path

~/.codex/config.toml

Enter your API key: Replace your-api-key in experimental_bearer_token = "your-api-key" with the actual key copied from the API Key page. Keep the double quotes, then save the configuration.

Important: You must use https://api.get-apikey.com/v1 as the API Base URL. Set base_url to this exact address, including /v1.

model = "gpt-5.5"
model_reasoning_effort = "high"
model_provider = "get-codex"

[model_providers.get-codex]
name = "Get-Codex"
base_url = "https://api.get-apikey.com/v1"
wire_api = "responses"
requires_openai_auth = true
experimental_bearer_token = "your-api-key"
4. Restart and test Codex CLI
Save the configuration, exit any running Codex CLI session, and reopen your terminal. From your project directory, run:
codex
Send hi. A normal response confirms that the CLI is configured.

FAQ

Got questions? Check here first

Which AI models are supported?

Get-Codex supports the latest gpt-5.5 xHigh and gpt-image-2 models.

How do I check usage and costs?

After logging in, go to the "Usage" page to view token consumption and cost breakdowns for each request.

Open Usage

Are API Keys secure?

API Keys are encrypted at rest and transmitted over HTTPS. We recommend rotating keys regularly and deleting unused ones.

How to handle request timeouts?

• Check network and proxy settings
• Verify upstream service status
• Long prompts may take more time — consider using streaming
• If timeouts persist, contact the admin to check upstream account health

Is WebSocket real-time communication supported?

Yes. Get-Codex supports OpenAI's WebSocket Mode for real-time conversations. Enable WebSocket support in the account settings.

Get Help

For setup, purchase, or usage issues, contact us by email or on WeChat Work.

WeChat Work

You can also scan the WeChat Work QR code for setup, purchase, and enterprise support.

WeChat Work QR CodeWeChat Work QR Code
getcodex001@gmail.com