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-Next-80B-A3B-Thinking 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-Next-80B-A3B-Thinking 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-Next-80B-A3B-Thinking",
    messages=[
      {
        "role": "user",
        "content": "Explain quantum computing in simple terms"
      }
    ],
    max_tokens=8192,
    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-Next-80B-A3B-Thinking is a next-generation foundation model from Alibaba’s Qwen team featuring a revolutionary Hybrid Attention mechanism (Gated DeltaNet + Gated Attention) with High-Sparsity MoE architecture.
  • With 80B total parameters and only 3.9B active per token, it delivers 10x higher throughput than Qwen3-32B on long contexts while outperforming Gemini-2.5-Flash-Thinking on multiple benchmarks.
  • The model operates in thinking-only mode for deep chain-of-thought reasoning, with a native 256K context window suited for complex multi-step tasks and long-horizon agentic planning.

​
Model at a Glance

FeatureDetails
Model IDQwen/Qwen3-Next-80B-A3B-Thinking
ProviderAlibaba Cloud (Qwen Team)
ArchitectureHybrid Attention (Gated DeltaNet + Gated Attention) with High-Sparsity MoE — 80B total / 3.9B active
Model Size80B Total / 3.9B Active
Parameters4
Context Length256K Tokens
Release Date2025
LicenseApache 2.0
Training DataLarge-scale multilingual dataset with RL post-training for deep chain-of-thought reasoning

​
When to use?

You should consider using Qwen3-Next-80B-A3B-Thinking if:
  • You need complex multi-step reasoning and mathematical proofs
  • Your application requires code synthesis and logical analysis
  • You are building agentic planning pipelines
  • Your use case involves long-context document analysis at high throughput
  • You need a thinking model that outperforms Gemini-2.5-Flash-Thinking
  • Your workflow requires a 256K context window with efficient sparse inference

​
Inference Parameters

Parameter NameTypeDefaultDescription
StreamingbooleantrueEnable streaming responses for real-time output.
Temperaturenumber0.6Controls randomness. Lower values recommended for reasoning tasks.
Max Tokensnumber8192Maximum number of tokens to generate.
Top Pnumber0.95Nucleus sampling parameter.

​
Key Features

  • Hybrid Attention Architecture: Combines Gated DeltaNet and Gated Attention layers for superior long-context efficiency over standard transformers.
  • 10x Throughput on Long Contexts: Delivers 10x higher throughput than Qwen3-32B on sequences of 32K+ tokens.
  • High-Sparsity MoE: Only 3.9B parameters active per token from 80B total, enabling frontier reasoning at low inference cost.
  • Native 256K Context Window: Supports long-horizon document analysis, multi-turn agentic tasks, and extended reasoning chains.
  • Outperforms Gemini-2.5-Flash-Thinking: Achieves higher benchmark scores than Gemini-2.5-Flash-Thinking across reasoning evaluations.
  • Apache 2.0 License: Fully open-source with unrestricted commercial use.

​
Summary

Qwen3-Next-80B-A3B-Thinking is Alibaba’s next-generation reasoning model built for high-throughput, deep chain-of-thought inference.
  • It uses a novel Hybrid Attention (Gated DeltaNet + Gated Attention) with High-Sparsity MoE, with 80B total and 3.9B active parameters per token.
  • It delivers 10x throughput over Qwen3-32B on long contexts and outperforms Gemini-2.5-Flash-Thinking on reasoning benchmarks.
  • The model supports a native 256K context window, thinking-only mode, and is optimized for complex reasoning, code synthesis, and agentic tasks.
  • Licensed under Apache 2.0 for full commercial use.