Qwen 3.5 397B
Access Qwen 3.5 397B through one API key. Alibaba's largest MoE model.
Alibaba's largest MoE model. 397B total parameters with 17B active, offering frontier-class reasoning.
No credit card required. Takes 30 seconds.
2,400+
Developers
1.2M+
API calls served
100+
Endpoints
$0.01
Per call
Yep, that's it.
Try it live
Send a message and see Qwen 3.5 397B respond in real time.
Maximum tokens in the response.
Real-time tokens
Context Window
262K tokens
Max Output
66K tokens
Input Price
$0.55 / 1M tokens
Output Price
$3.28 / 1M tokens
Strengths
Qwen 3.5 397B is the largest model in Alibaba's Qwen 3.5 family, offering the most capacity for complex reasoning and generation.
With 397 billion total parameters and 17 billion active per token, it combines frontier-scale capacity with mixture-of-experts efficiency.
Delivers frontier-class reasoning across math, coding, and multi-step problem solving, the top tier of the Qwen lineup.
Its mixture-of-experts design activates only 17B of 397B parameters per token, keeping inference far cheaper than a dense model of equivalent capacity.
Quick start
Copy this snippet and start making calls with Qwen 3.5 397B.
const res = await fetch('https://api.yepapi.com/v1/ai/chat', {
method: 'POST',
headers: {
'x-api-key': 'YOUR_API_KEY',
'Content-Type': 'application/json',
},
body: JSON.stringify({
"model": "qwen/qwen3.5-397b-a17b",
"messages": [
{
"role": "user",
"content": "Explain API gateways in 2 sentences."
}
],
"maxTokens": 256
}),
});
const { data } = await res.json();
console.log(data.message.content);Why use Qwen 3.5 397B through YepAPI?
Qwen 3.5 397B API: Alibaba's largest MoE model for frontier reasoning
Qwen 3.5 397B is Alibaba's largest mixture-of-experts model, with 397 billion total parameters and 17 billion active per token. It delivers frontier-class reasoning across math, code, and complex problem solving while keeping inference efficient.
Through YepAPI you call Qwen 3.5 397B with one API key on the chat endpoint at $0.55 per 1M input and $3.28 per 1M output tokens, with a 262K context window and 65K output.
What is Qwen 3.5 397B?
Qwen 3.5 397B is the flagship of Alibaba's Qwen 3.5 family — its largest model. It uses a mixture-of-experts architecture with 397 billion total parameters but only about 17 billion active per token, so it reaches frontier-scale capability while running far more efficiently than a dense model of similar size. It excels at demanding reasoning: advanced mathematics, complex code generation, and multi-step problem solving. With a 262K-token context window and up to 65,536 tokens of output, it handles large inputs and long, detailed responses, making it the choice when you need the strongest reasoning the Qwen lineup offers.
Build with Qwen 3.5 397B via YepAPI
Send chat completion requests to Qwen 3.5 397B over a single YepAPI key on the /v1/ai/chat endpoint — no Alibaba Cloud setup needed. Pass your messages, optionally stream the output, and let the MoE routing run server-side. Because one key reaches the whole catalog, you can route your hardest reasoning and coding tasks to Qwen 3.5 397B and fall back to cheaper models for routine work, all without re-authenticating. It drops cleanly into agentic systems, coding assistants, and analysis pipelines.
Qwen 3.5 397B API pricing — $0.55/1M input, $3.28/1M output
Qwen 3.5 397B costs $0.55 per 1M input tokens and $3.28 per 1M output tokens on YepAPI. Thanks to MoE sparsity, you get frontier-class reasoning at a price well below dense models of comparable capability, so reserving it for your toughest prompts is economical. You pay only for tokens used, with no minimums, and one balance covers it alongside every other model on the platform.
Qwen 3.5 397B for complex reasoning and code
Qwen 3.5 397B is the model to deploy on hard problems: intricate algorithm design, advanced math, large refactors, and multi-step agentic reasoning where smaller models fall short. Its 262K context lets it hold substantial code or documents while it reasons, and its 65K output supports long, detailed solutions. When correctness on difficult tasks matters most, its frontier-class capability earns the higher per-token cost.
Try Qwen 3.5 397B free
Get $5 in free credit with no card required. Sign up, grab your API key, and put Qwen 3.5 397B on a hard reasoning or coding task to gauge its frontier-class quality before committing.
Start generating in 30 seconds
$5 free credit on signup. No credit card required. Pay per call.
What developers say
“Switched from SerpAPI and cut our SERP costs by 80%. Same data quality, way simpler billing.”
“One API key for AI models, SERP data, and web scraping. Saved us from managing 4 separate providers.”
“The $5 free credit let us prototype our entire rank tracking feature before committing. No other API does that.”
Frequently asked questions
Alibaba's largest MoE model. 397B total parameters with 17B active, offering frontier-class reasoning.
Input tokens cost $0.55 per 1M tokens and output tokens cost $3.28 per 1M tokens through YepAPI. No monthly minimums — you only pay for what you use.
Sign up for a free API key, then send requests to the /v1/ai/chat endpoint.
Qwen 3.5 397B supports a 262K token context window with up to 66K output tokens per request.
Ready to use Qwen 3.5 397B?
$5 free credit on signup. No credit card required. Pay per call.
Explore more models
Qwen3.7 Max
QwenAccess Qwen3.7 Max through one API key. Alibaba's most capable Qwen model.
Qwen 3.6 Plus
QwenAccess Qwen 3.6 Plus through one API key. Powerful reasoning with 1M context at a great price.
Qwen 3 Coder
QwenAccess Qwen 3 Coder through one API key. Coding-specialized model at an ultra-low price.