Cursor

Using Cursor with Fast-Token

Cursor lets you use your own API key so you can spend purchased quota (for example DeepSeek or a relay service) instead of being limited by Cursor’s official subscription usage caps.

Step 1: Enable API mode

1. Open Cursor.
2. Open settings:
   • Click the gear icon (Cursor Settings) in the top-right corner.
   • Or use Ctrl + Shift + J (Windows/Linux) or Cmd + Shift + J (macOS).

3. In the sidebar or window that opens, select the General or Models tab.
4. Find OpenAI API Key and paste an API key from the Fast-Token platform.
5. Turn on Override OpenAI Base URL and set it to https://fast-token.com/v1.

Step 2: Manage the model list (Models)

To make sure Cursor calls the right models, enable or add them manually in settings. Custom model IDs must not match Cursor’s built-in model names. Choose models from the Cursor category in the model list.

1. On the same settings page, find the Models section.

2. Turn on the switch for the models you want (for example gpt-4o, claude-3-5-sonnet).

3. Add a custom model: for DeepSeek and similar providers, click + Add Model.

4. Enter the exact model ID (for example deepseek-chat or deepseek-reasoner).

Step 3: Use in the editor

1. Return to the editor.
2. Use Ctrl + L (Chat) or Ctrl + K (Edit).
3. In the model dropdown at the bottom of the input area, select the third-party model you configured.
4. Note: When using your own API key, some advanced Cursor features (such as certain Composer optimizations) may behave differently; code generation quality depends on the API you choose.

FAQ

Symptom	Fix
Verify Failed	Make sure the Base URL does not include /chat/completions — use only https://fast-token.com/v1.
Model does not respond	Check that your Fast-Token account has sufficient balance.
Cannot use built-in models	After enabling your own API key, Cursor usually spends your key quota before your subscription quota.