Want to use DeepSeek-V4 or Qwen3 in Cursor IDE instead of expensive GPT-4? This guide shows you how to configure Cursor with Chinese LLMs through AI Token gateway — saving up to 95% on your AI coding costs.
Cursor Pro costs $20/month for limited GPT-4 usage. With Chinese LLMs through AI Token, you get unlimited access to models that are 90-95% cheaper with comparable code quality.
Step 1: Get Your API Key
Sign up at AI Token Gateway and get your API key. Free credits included.
Step 2: Open Cursor Settings
In Cursor, go to Settings → Models (or Cursor Settings → OpenAI Key)
Step 3: Configure OpenAI Override
Enable "Override OpenAI Base URL" and enter:
Base URL: https://eaf9553505eeb8f5-115-190-107-107.serveousercontent.com/v1 API Key: your-aitoken-api-key
Step 4: Select Your Model
Choose from available models:
Make sure you're using the full base URL including /v1 at the end. The format should be: https://eaf9553505eeb8f5-115-190-107-107.serveousercontent.com/v1
Use the exact model name from our model list. Model names are case-sensitive and use hyphens (e.g., deepseek-v4, not DeepSeek-V4).
Try using Flash models (deepseek-v4-flash, qwen3-flash, glm-4-flash) for faster responses. They're also cheaper!