Home / Model Catalog
One API key for all leading Chinese LLMs. Every model uses the OpenAI-compatible format, so switching is just a Base URL change.
The table shows indicative converted pay-as-you-go unit prices. Subscribers are charged in-plan rates, which are usually lower than the prices below.
| Vendor | Representative models | Context | Input (per 1M tokens) | Output (per 1M tokens) | Status |
|---|---|---|---|---|---|
| DeepSeek | DeepSeek-V3 / R1 | 128K | from ¥2 | from ¥8 | Available |
| Qwen | Qwen-Max / Plus | 128K | from ¥4 | from ¥12 | Available |
| Kimi | Moonshot series | 200K | from ¥4 | from ¥12 | Available |
| Doubao | Doubao-Pro | 256K | from ¥0.8 | from ¥2 | Available |
| Zhipu AI | GLM-4 series | 128K | from ¥5 | from ¥5 | Available |
| Baichuan | Baichuan series | 32K | from ¥2 | from ¥4 | Available |
| iFlytek Spark | Spark series | 32K | from ¥3 | from ¥6 | Available |
| MiniMax | MiniMax-Text | 200K | from ¥1 | from ¥8 | Available |
| StepFun | Step series | 128K | from ¥3 | from ¥12 | Available |
Prices shown are indicative ranges. Actual unit prices follow upstream vendor list prices and our live pricing; exact figures are available in the console's model catalog.
OpenAI-compatible format: migrate existing projects by changing the Base URL.
from openai import OpenAI
client = OpenAI(
base_url="https://api.tokenhub.cn/v1",
api_key="sk-YOUR_API_KEY"
)
resp = client.chat.completions.create(
model="deepseek-v3",
messages=[{"role": "user", "content": "Hello"}]
)
print(resp.choices[0].message.content)
curl https://api.tokenhub.cn/v1/chat/completions \
-H "Authorization: Bearer sk-YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek-v3",
"messages": [{"role":"user","content":"Hello"}],
"stream": true
}'
Sign up and receive a free 3-day trial subscription with ¥10 of usable credit, valid for all models listed above.