Integration steps
- Create a restricted server-side key.
- Set the compatible base URL.
- Select a Qwen model from the live catalog.
- Run a basic request and verify usage.
- Test tools or streaming only when the model advertises support.
What this page does not assume
It does not assume every Qwen release is enabled, or that every model supports the same feature set. The catalog below is filtered from the active public list.
qwen.pypython
from openai import OpenAI
client = OpenAI(
api_key="qh_live_...",
base_url="http://localhost:4000/v1",
)
response = client.chat.completions.create(
model="qwen3.7-plus",
messages=[{"role": "user", "content": "Explain this error log."}],
)
print(response.choices[0].message.content)Live model catalog
| Model | Input / 1M tokens | Output / 1M tokens | Context | Status |
|---|---|---|---|---|
| qwen-vl-plus | CNY 1 | CNY 2.5 | 131K | Available |
| qwen3.5-flash | CNY 0.25 | CNY 2.4 | 1M | Available |
| qwen3.6-flash | CNY 1.45 | CNY 8.65 | 1M | Available |
| qwen3.7-plus | CNY 2.4 | CNY 9.6 | 1M | Available |
| qwen3.7-max | CNY 14.4 | CNY 43.2 | 1M | Available |
Common questions
Can I use the OpenAI Python client with Qwen?
Yes, for the OpenAI-compatible chat completions path. Set the gateway base URL, your Qingheng API key and a public Qwen model name from the catalog.
Does every Qwen model support function calling?
No assumption is made. Check the tools capability in the live catalog and run a schema-validation test before production use.
How do I see current Qwen prices?
Use the Qwen pricing page or the model table below. Both read the current public catalog.