Requesty
DeepInfra Inc. logo
DeepInfra Inc.

Qwen/Qwen3.5-397B-A17B

Qwen3.5-397B-A17B is Alibaba's most capable Qwen3.5 model, a Mixture-of-Experts architecture with 397B total parameters and 17B activated per token. It features a 262K token context window (extensible to 1M with YaRN), thinking/reasoning mode, tool calling with MCP integration, and support for 201 languages. Sets state-of-the-art results on reasoning, coding, math, and multimodal benchmarks.

VisionReasoningTool callingCachingJSON schema
CompareDocs

Specifications

Context window262K tokens
Max outputN/A
API typechat
AddedMay 27, 2026
Model IDdeepinfra/Qwen/Qwen3.5-397B-A17B
Data retentionNo
Used for trainingNo
Provider location🇺🇸 US
Privacy policyDeepInfra Privacy Policy→

Benchmarks

Released 2026-02-16
Coding Indexcoding
48.2%

Artificial Analysis Coding Index — a composite of coding evaluations including LiveCodeBench, SciCode and Terminal-Bench.

GPQA Diamondreasoning
89.3%

Graduate-level physics, chemistry & biology questions designed to resist Googling.

Intelligence Indexreasoning
33.7%

Artificial Analysis Intelligence Index — a composite of multiple evaluations measuring overall model capability.

Scores are sourced from official model cards, Artificial Analysis, and public leaderboards. Benchmarks measure specific skills and do not capture every aspect of model quality. Always test on your own workload.

Pricing

Prices updated July 24, 2026
Input / 1M
$0.49
Output / 1M
$3.60
Cache write
N/A
Cache read / 1M
$0.30
Estimated cost
100K input + 10K output$0.0850
1M input + 100K output$0.85
10M input + 1M output$8.50

Requesty charges exactly what the upstream provider charges, no markup, no per-request fees. Prompt caching and smart routing can reduce effective cost by 30-80%.

Quickstart

Drop-in compatible with the OpenAI SDK. Change the base URL, swap in your Requesty API key, and set the model to deepinfra/Qwen/Qwen3.5-397B-A17B.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_REQUESTY_API_KEY",
    base_url="https://router.requesty.ai/v1",
)

response = client.chat.completions.create(
    model="deepinfra/Qwen/Qwen3.5-397B-A17B",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

print(response.choices[0].message.content)

Other DeepInfra Inc. models

zai-org/GLM-5.1
203K ctx$1.05/1M
Coding 56
nvidia/Nemotron-3-Nano-30B-A3B
262K ctx$0.05/1M
nvidia/NVIDIA-Nemotron-3-Super-120B-A12B
262K ctx$0.10/1M
Coding 38
moonshotai/Kimi-K2.6
262K ctx$0.75/1M
Coding 62
gemma-4-31B-it
262K ctx$0.13/1M
deepseek-ai/DeepSeek-V4-Pro
1.0M ctx$1.30/1M
Coding 59

Frequently asked questions

How much does Qwen/Qwen3.5-397B-A17B cost?
Qwen/Qwen3.5-397B-A17B is priced at $0.49 per million input tokens and $3.60 per million output tokens when accessed via Requesty. Prompt caching is supported, which can cut effective input cost by up to 90% on repeated context. Requesty charges exactly what the upstream provider charges, we don't add markup.
What is the context window of Qwen/Qwen3.5-397B-A17B?
Qwen/Qwen3.5-397B-A17B has a context window of 262K tokens. That's roughly 350 words of input you can fit in a single prompt.
How does Qwen/Qwen3.5-397B-A17B perform on benchmarks?
Qwen/Qwen3.5-397B-A17B scores 95.6% on τ²-Bench, 89.3% on GPQA Diamond, 48.2% on Coding Index. See the full benchmark chart above for results across MMLU Pro, GPQA Diamond, SWE-Bench Verified, HumanEval, MATH, AIME, MMMU, and LiveBench.
What can Qwen/Qwen3.5-397B-A17B do?
Qwen/Qwen3.5-397B-A17B supports vision input, tool calling, extended reasoning, prompt caching, structured outputs (JSON schema). You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use Qwen/Qwen3.5-397B-A17B with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "deepinfra/Qwen/Qwen3.5-397B-A17B". The Quickstart above shows Python, JavaScript and cURL snippets.
Can I run Qwen/Qwen3.5-397B-A17B through Requesty?
Yes. Qwen/Qwen3.5-397B-A17B runs through Requesty's OpenAI-compatible API, served from DeepInfra Inc.. You do not host the model yourself: point base_url at Requesty, set the model to "deepinfra/Qwen/Qwen3.5-397B-A17B", and requests are routed to the upstream provider with automatic failover. The same key gives you 600+ other models too.

Access Qwen/Qwen3.5-397B-A17B through Requesty

One API key, 600+ models, OpenAI-compatible. No markup on provider prices, automatic failover, and smart caching built-in.

All DeepInfra Inc. models