Using Cursor with Yi-AI Router

Connect Cursor through Yi-AI Router to access all models via a single API key, with centralized billing and automatic provider failover.

Quick Start

Prerequisites

Cursor installed
A Yi-AI account with an API Key (sk-yi-...)

Configuration

Cursor supports custom OpenAI-compatible endpoints through its Override OpenAI Base URL feature.

Step 1: Open Cursor Settings

Open Cursor Settings (Cmd/Ctrl + ,), navigate to Models, then expand the API Keys section.

Step 2: Configure OpenAI Provider

Enable OpenAI API Key and fill in:

Field	Value
OpenAI API Key	Your Yi-AI API Key (`sk-yi-...`)
Override OpenAI Base URL	`https://www.yiairouter.com/api/v1`

Step 3: Add Models

In the Models section, click + Add Model and enter model IDs from the Yi-AI catalog. For example:

claude-sonnet-4-20250514
claude-haiku-4-5
gpt-4o-mini
gemini-1.5-flash

You can browse all available models at your Yi-AI Dashboard or via the API at /api/v1/models.

Step 4: Select Model

Open the model picker in Cursor's chat or Composer panel and select one of the models you added. All requests will now route through Yi-AI Router.

How It Works

Yi-AI Router implements the OpenAI Chat Completions API (/api/v1/chat/completions). Cursor's OpenAI provider slot communicates with Yi-AI Router directly using its native protocol — no local proxy required. The platform handles model mapping, usage accounting, and billing.

Key features:

Single API key for all models across providers
Automatic model routing — deprecated models fall back to stable alternatives
Centralized billing — all usage tracked in your Yi-AI dashboard
No vendor lock-in — switch models without changing configuration

Model Discovery

Cursor can auto-discover available models from the /api/v1/models endpoint. After configuring the OpenAI provider, Cursor may automatically populate the model list. If not, add models manually using the IDs shown in the Yi-AI dashboard.

Troubleshooting

"Invalid API Key"
Make sure you're using your Yi-AI API key (sk-yi-...), not an OpenAI or other provider key.

Models not appearing
Verify your Yi-AI account has sufficient balance. Models with $0 balance may be restricted.

Request timeouts
If using a local development server (localhost), ensure the server is running. For the production gateway, check your network connection.

Model not found
Ensure the model ID matches exactly what's shown in the Yi-AI model catalog. Model IDs are case-sensitive.