J JJAPI
δΈ­ζ–‡

Alibaba

Use the Qwen API without an Alibaba Cloud account.

Qwen-Max, Qwen-Plus, Qwen-Turbo, Qwen-VL, and Qwen-Coder β€” China's strongest open-weight family, accessible through one OpenAI-compatible endpoint.

Get my API key β†’ Quickstart in 30 seconds β†’

Latest models as of 2026-05-23

See all models β†’
qwen3-max flagship

Alibaba's strongest open-weight model. Particularly strong on Chinese benchmarks.
qwen3-plus balanced

Production-grade balance. The default Qwen for most apps.
qwen3-turbo fast

Fast cheap variant. For high-volume Chinese-language traffic.

Why use JJAPI for this API

Best-in-class for Chinese-language tasks

Qwen consistently leads benchmarks for Chinese reasoning, translation, and summarization.

No Alibaba Cloud setup

Skip Aliyun account creation, real-name verification, and the DashScope dashboard.

Open-weight compatibility

If you switch from Qwen-API to a self-hosted Qwen later, the prompt format is identical.

OpenAI-compatible interface

Same SDK, same request shape β€” no Qwen-specific quirks.

example.py 
from openai import OpenAI

client = OpenAI(
    base_url="https://api.jjapi.net/v1",
    api_key="sk-jjapi-...",
)

r = client.chat.completions.create(
    model="qwen3-max",  # current as of 2026-05-23
    messages=[{"role": "user", "content": "Hi"}],
)
print(r.choices[0].message.content)

Frequently asked

Can I use Qwen for Chinese-first apps?

Yes β€” Qwen is one of the best models for Chinese tasks. It significantly outperforms Western models on idiom comprehension and code-switching.

Does Qwen support vision?

qwen-vl-max accepts image_url and base64 image inputs in OpenAI vision format.

Is Qwen-Coder faster than GPT-4o for code?

On code-specific benchmarks Qwen-Coder is roughly equivalent to GPT-4o at a fraction of the price.

Get my API key β†’