Alibaba's Best AI Models

Qwen API Access — All Qwen Models, One Gateway

Access the full Qwen model family through a single OpenAI-compatible API. No need for separate Alibaba Cloud accounts — just one API key for all Qwen models at the best available prices.

Available Qwen Models & Pricing

Model	Input $/1M	Output $/1M	Context	Best For
Qwen3-Flash	$0.05	$0.10	128K	Fast, balanced tasks
Qwen-Plus	$0.15	$0.30	128K	General purpose
Qwen-Max	$0.40	$0.80	128K	Complex reasoning

Why Qwen?

Alibaba's Qwen models are among the best Chinese-origin LLMs available today:

Top-tier multilingual: Excellent in Chinese, English, and 20+ languages
Strong coding: Competitive with GPT-4 on code generation benchmarks
Long context: 128K tokens standard across all models
Open ecosystem: Qwen3 is open-weight, enabling broad adoption
Fast inference: Qwen3-Flash delivers 100+ tokens/sec

Qwen vs. Competitors

Metric	Qwen-Max	DeepSeek-V4	GPT-4o
Input cost/1M tokens	$0.40	$0.28	$2.50
MMLU Score	86%	88%	88%
HumanEval (coding)	85%	89%	90%
Chinese NLP	★★★★★	★★★★★	★★★☆☆
Context window	128K	128K	128K

Quick Start

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://eaf9553505eeb8f5-115-190-107-107.serveousercontent.com/v1"
)

response = client.chat.completions.create(
    model="qwen3-flash",  # or qwen-max, qwen-plus
    messages=[
        {"role": "user", "content": "分析这段代码的性能瓶颈"}
    ]
)
print(response.choices[0].message.content)

Access Qwen Models Now →

← Back to AI Token Gateway