Base URL
https://dash.chinaapi.ai/v1. Chat completions are available at /v1/chat/completions.
Tool integration guides
Use Chinese AI models in the tools you already run.
ChinaAPI gives international developers one OpenAI-compatible endpoint for DeepSeek, Kimi, Qwen, GLM, and MiniMax models, billed in USD at official China list prices.
Setup tutorials
Five OpenAI-compatible integration paths.
Cursor with Chinese modelsConfigure Cursor's OpenAI API key and custom base URL for DeepSeek, Kimi, Qwen, and GLM coding workflows.
Cline with Chinese modelsUse the OpenAI Compatible provider in the VS Code agent with ChinaAPI model IDs.
LiteLLM with Chinese modelsRoute LiteLLM proxy and SDK calls through ChinaAPI using
openai/<model> entries and api_base.
LobeChat with Chinese modelsSet the OpenAI provider endpoint to ChinaAPI for chat access to Chinese LLM families.
n8n with Chinese modelsCreate OpenAI credentials with a custom base URL and use them in AI Agent or OpenAI LLM nodes.
Connection details
Same endpoint across tools.
API key
Register, open Dashboard, then go to Console -> Tokens. Keys use the sk-... format.
Trial
Signup includes $1 free credit for self-serve testing.
Pricing
Check live pricing before production runs. Model availability and IDs can change.
Start testing
Create one key and reuse it across the guides.
Use the same ChinaAPI token in Cursor, Cline, LiteLLM, LobeChat, n8n, or direct OpenAI-compatible SDK calls.
Good first models
deepseek-v4-flash, qwen3.6-flash, and MiniMax-M2.7 are sensible low-latency first tests.
Heavier reasoning
Try qwen3.7-max, glm-5.2, deepseek-v4-pro, or kimi-k2.7-code when output quality matters more than speed.