China's best AI models.
One simple API.
DeepSeek, Qwen and GLM — served through a single OpenAI-compatible endpoint. Built for developers and teams outside China who want frontier Chinese models without the setup friction.
https://api.chinamodels.io/v1
Frontier models, ready to call
Swap the model name — everything else works exactly like the OpenAI API you already know.
DeepSeek
Strong general reasoning and coding at a famously efficient price point. Includes the R1-series reasoning model.
Qwen
Alibaba's flagship model family — broad multilingual coverage and strong tool-use, from turbo tiers to Max.
GLM
Zhipu's GLM series — capable bilingual models with competitive long-context performance.
Ship in under a minute
- ⚡OpenAI-compatible Point your existing SDK at our base URL. No new client library, no rewrites.
- 🔑One key, every model A single API key routes to all supported providers. We handle the upstream accounts.
- 🔒TLS everywhere Traffic is encrypted end-to-end and served through Cloudflare's global edge.
- 📊Usage dashboard Track spend per token, per model and per key in the console.
curl https://api.chinamodels.io/v1/chat/completions \ -H "Authorization: Bearer $YOUR_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "deepseek-chat", "messages": [{"role":"user","content":"Hello!"}] }' # → {"choices":[{"message":{"content":"Hi there! ..."}}]}
Three steps to production
Create an account
Sign up in the console and generate your API key — no Chinese phone number or local billing required.
Point your SDK
Set the base URL to api.chinamodels.io/v1 and pick a model. Works with any OpenAI-compatible client.
Scale as you go
Pay-as-you-go metering per token. Add more models and keys for your team whenever you need.
Start building with Chinese frontier models
Get an API key and make your first call in minutes.
Open the console →