Integrations

Since DeepToken is 100% compatible with the OpenAI API specification, you can connect it to any third-party client, editor, or platform that supports custom OpenAI endpoints.

You only need two configurations:

Base URL: https://api.deeptoken.app/v1
API Key: Your DeepToken API key (dtk_...)

Developer Tools

Select the tool you want to integrate with DeepToken below:

Cursor Configuration

To configure DeepToken in Cursor:

Open Cursor and go to Settings (gear icon in the top right) -> Cursor Settings.
Click on the Models tab in the sidebar.
Scroll down to the OpenAI API Key section.
Click Configure override (or edit button) next to the API key input.
Set the following:
- OpenAI API Key: <YOUR_DEEPTOKEN_API_KEY>
- Base URL Override: https://api.deeptoken.app/v1
Under Models, toggle off default models you do not use and add the custom model IDs you wish to call from your DeepToken catalog (e.g. gpt-4o-mini, claude-3-5-sonnet).

Mẹo

DeepToken automatically handles failovers and routes calls to the most optimal upstream channel. You only need to add the canonical model names in Cursor settings.

Cline Configuration

To configure DeepToken in Cline (formerly Devins):

Open VS Code and open the Cline panel.
Click on the Settings icon (gear) in the Cline panel.
Under API Provider, select OpenAI Compatible.
Configure the settings as follows:
- Base URL: https://api.deeptoken.app/v1
- API Key: <YOUR_DEEPTOKEN_API_KEY>
- Model ID: Enter the model you want to use (e.g. claude-3-5-sonnet or gpt-4o-mini).
Click Save to start coding with DeepToken.

Desktop Chat Clients

For applications like Cherry Studio or Chatbox:

Open the application settings and navigate to API Providers / Models.
Add a new Custom OpenAI or OpenAI Compatible provider.
Name it DeepToken.
Enter the credentials:
- API Key: <YOUR_DEEPTOKEN_API_KEY>
- Base URL / API Host: https://api.deeptoken.app/v1 (Note: some clients need /chat/completions stripped, others require /v1 suffix).
Fetch or manually add the list of models you wish to use.