All models

QwenAlibaba CloudOpen weights

Qwen 3.5 397B A17B.

Qwen 3.5 397B A17B is the native vision-language model of the Qwen 3.5 series, with a hybrid architecture of linear attention and mixture-of-experts that activates 17 billion of its 397 billion parameters per token. It covers language understanding, reasoning, code generation, agent tasks, image and video understanding and the operation of graphical interfaces, and processes text, images and video with 262K context.

Get started → Providers and prices

base_url="https://api.noirdoc.de/v1" model="Qwen/Qwen3.5-397B-A17B"

  • Data centres in Germany
  • § 203 confidentiality
  • Zero retention
  • List prices

Details.

Where the model runs, which rules apply and what it can do.

  1. Data centres in Germany IONOS Contracting party: IONOS Cloud GmbH · based in Montabaur, Germany
§ 203 German Criminal Code (StGB)
Confidentiality agreed with IONOS
Weights
Open weights
Features
Tools, Vision, Reasoning and JSON
See how this looks in your account →
Map: Qwen 3.5 397B A17B runs in data centres in Germany (IONOS).

Providers.

We offer Qwen 3.5 397B A17B through these providers.

Provider § 203 model=
€ · 1M tokens Input Cache Read Cache Write Output
IONOS DE German data centres
§ 203 agreed
Qwen/Qwen3.5-397B-A17B
Input 0.60 Output 3.60

More models.

Alibaba Cloud's Qwen series covers open models for chat, reasoning, code, vision and embeddings in many sizes.

All Qwen models →
Model
€ · 1M tokens Input Cache Read Cache Write Output
Qwen 3.8 27B Tools Vision Reasoning JSON Compact Qwen model with 27B parameters — a balanced generalist for chat, reasoning and tool calls. weber.cloud Logo weber.cloud · DE AKI.IO Logo AKI.IO · DE
Input from 0.15 Cache Read 0.10 Output 0.65
Qwen 3.6 35B A3B Tools Vision Reasoning JSON Efficient Qwen MoE — strong in multilingual work and reasoning at low cost. AKI.IO Logo AKI.IO · DE
Input 0.15 Cache Read — Output 0.50
Qwen 3.5 9B Tools Vision Reasoning JSON Compact multimodal Qwen model with 9B parameters — an efficient architecture for reasoning, code and image understanding. Regolo Logo Regolo · EU
Input 0.10 Cache Read — Output 0.20
Qwen3-Coder-Next Code Tools JSON Specialised code model for generation, refactoring and agentic dev workflows. IONOS Logo IONOS · DE
Input 0.15 Cache Read — Output 0.80
Qwen3-VL Embedding 8B Embeddings Large multimodal embedding model for demanding search and RAG applications. IONOS Logo IONOS · DE
0.11 € per 1M tokens
Qwen3-VL Reranker 8B Reranking Reranker for more precise hits in RAG pipelines — reorders search results. IONOS Logo IONOS · DE
0.045 € per 1M tokens

Get started.

Test selected models right away with €5 starter credit — every model from the first top-up. If you have questions first, talk to our sales team.

Sign in →
python · openai sdk api.noirdoc.de/v1 
import os
from openai import OpenAI

client = OpenAI(
    # one line, that's all
    base_url="https://api.noirdoc.de/v1",
    api_key=os.environ["NOIRDOC_API_KEY"],
)

response = client.chat.completions.create(
    model="Qwen/Qwen3.5-397B-A17B",  # provider & model
    messages=[{"role": "user", "content": "Hello"}],
)