Use with Cursor

Cursor lets you enter a custom API Key in its settings. Setting names vary slightly between versions — follow your current version's UI.

Prepare an API Key

Create a Cursor-specific API Key in the yylx.io console first. A name like cursor-desktop or cursor-work makes it easy to spot in the usage log later.

Cursor usually goes through the OpenAI-compatible configuration, so you need three things:

Setup steps

1. Open Cursor's settings.
2. Go to the Models or API Keys section.
3. Find the OpenAI API Key, custom OpenAI-compatible configuration, or Override OpenAI Base URL setting.
4. Enter the Key created in the yylx.io console as the API Key.
5. For Base URL or Override OpenAI Base URL, enter:

https://app.yylx.io/v1

1. If the UI allows adding models manually, enter a model name shown in the yylx.io console.
2. Save, then click Verify or start a new chat to test.

Model names

If Cursor allows adding models manually, use the model names shown in the yylx.io console.

If your Cursor version doesn't support a custom Base URL, it can't point at the yylx.io gateway directly. Consider Codex, Claude Code, or Cherry Studio instead, or wait for a Cursor version that supports the setting.

Testing

After saving the configuration, start with a very lightweight test:

Reply with a single sentence: Cursor is connected.

If it responds normally, go ahead with Composer, Chat, or Agent. If the test fails, don't keep retrying big tasks — check the Key, Base URL, and model name first.

Notes

Cursor's custom API Key usually covers only the standard chat-model capabilities. Tab Completion, built-in models, and some Agent features may still go through Cursor's own service.

If Verify fails, check: