Skip to main content

​
About the Provider

Alibaba Cloud is the cloud computing arm of Alibaba Group and the creator of the Qwen model family. Through its open-source initiative, Alibaba has released state-of-the-art language and multimodal models under permissive licenses, enabling developers and enterprises to build powerful AI applications across diverse domains and languages.

​
Model Quickstart

This section helps you quickly get started with the Qwen/Qwen3.5-397B-A17B model on the Qubrid AI inferencing platform. To use this model, you need:
  • A valid Qubrid API key
  • Access to the Qubrid inference API
  • Basic knowledge of making API requests in your preferred language
Once authenticated with your API key, you can send inference requests to the Qwen/Qwen3.5-397B-A17B model and receive responses based on your input prompts. Below are example placeholders showing how the model can be accessed using different programming environments.
You can choose the one that best fits your workflow. 
from openai import OpenAI

# Initialize the OpenAI client with Qubrid base URL
client = OpenAI(
    base_url="https://platform.qubrid.com/v1",
    api_key="QUBRID_API_KEY",
)

# Create a streaming chat completion
stream = client.chat.completions.create(
    model="Qwen/Qwen3.5-397B-A17B",
    messages=[
      {
        "role": "user",
        "content": [
          {
            "type": "text",
            "text": "What is in this image? Describe the main elements."
          },
          {
            "type": "image_url",
            "image_url": {
              "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
            }
          }
        ]
      }
    ],
    max_tokens=16384,
    temperature=0.6,
    top_p=0.95,
    stream=True
)

# If stream = False comment this out
for chunk in stream:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)
print("\n")

# If stream = True comment this out
print(stream.choices[0].message.content)

​
Model Overview

Qwen3.5-397B-A17B is Alibaba’s flagship open-source model and the first in the Qwen3.5 series, released February 16, 2026.
  • It is a native multimodal model trained from scratch on trillions of text, image, and video tokens using early fusion across 201 languages.
  • With 397B total parameters and 17B active per token, it outperforms all Qwen3-VL models on vision tasks while matching or exceeding text-only frontiers.
  • The hosted version is called Qwen3.5-Plus, supporting up to 1M token context.

​
Model at a Glance

FeatureDetails
Model IDQwen/Qwen3.5-397B-A17B
ProviderAlibaba Cloud (Qwen Team)
ArchitectureHybrid Gated DeltaNet + Sparse MoE Transformer — 60 layers (15 cycles of 3× DeltaNet + 1× Gated Attention), hidden size 4096, 248,320 vocab size
Model Size397B Total / 17B Active
Context Length256K Tokens (up to 1M via Qwen3.5-Plus API)
Release DateFebruary 16, 2026
LicenseApache 2.0
Training DataTrillions of multimodal tokens (text, image, video) across 201 languages and dialects; large-scale RL post-training across million-agent environments

​
When to use?

You should consider using Qwen3.5-397B-A17B if:
  • You need native multimodal reasoning across text, image, and video
  • Your application requires frontier-level agentic workflows and multi-tool orchestration
  • You are working on long-horizon code generation and system design
  • You need scientific research and mathematical problem solving
  • Your use case involves complex document understanding and RAG
  • You need GUI and web automation
  • Your application requires multilingual support across 201 languages

​
Inference Parameters

Parameter NameTypeDefaultDescription
StreamingbooleantrueEnable streaming responses for real-time output.
Temperaturenumber0.6Use 0.6 for non-thinking tasks, 1.0 for thinking/reasoning tasks.
Max Tokensnumber16384Maximum tokens to generate. Use higher values for thinking mode.
Top Pnumber0.95Nucleus sampling parameter.
Top Knumber20Limits token sampling to top-k candidates.
Enable ThinkingbooleanfalseToggle chain-of-thought reasoning mode. Set temperature=1.0 when enabled.

​
Key Features

  • Native Multimodal: First open-source model trained from scratch with early fusion of text, image, and video — no separate vision encoder.
  • Frontier Intelligence at Efficient Cost: 397B total / 17B active parameters via Sparse MoE for frontier-level performance at efficient compute.
  • 256K Native Context: Supports up to 256K tokens natively, extendable to 1M via the hosted Qwen3.5-Plus API.
  • Thinking Mode: Configurable chain-of-thought reasoning with enable_thinking=true for complex multi-step tasks.
  • Multi-Token Prediction (MTP): Enhanced throughput via MTP for faster inference.
  • 201 Language Support: Trained on 201 languages and dialects for broad multilingual coverage.
  • Apache 2.0 License: Full commercial freedom with open weights.

​
Summary

Qwen3.5-397B-A17B is Alibaba’s flagship open-source native multimodal model built for frontier-level intelligence across text, image, and video.
  • It uses a hybrid Gated DeltaNet + Sparse MoE architecture with 397B total and 17B active parameters.
  • The model supports 256K native context, configurable thinking mode, and 201 languages.
  • It achieves 87.8% on MMLU-Pro and outperforms all Qwen3-VL models on vision tasks.
  • Licensed under Apache 2.0 for full commercial use.