Qwen 3.6

Alibaba Cloud's flagship AI series featuring the powerful Qwen3.6-Max and ultra-efficient MoE models for industry-leading multilingual reasoning.

QwenAlibabaOpen SourceLanguage ModelLarge Language ModelAI AssistantMultilingualMoELatest
Developer
Alibaba Cloud
Type
Hybrid Language Model
License
Open Source (Apache 2.0)
🌐 Official Website💰 Pricing & API

Overview

Qwen 3.6, released by Alibaba Cloud on April 20, 2026, represents the cutting edge of the Tongyi Qianwen series. The flagship Qwen3.6-Max-Preview marks a strategic shift for Alibaba toward proprietary, frontier-class intelligence. Unlike its open-source predecessors, Qwen3.6-Max is a hosted model optimized for extreme reasoning, featuring an "always-on" chain-of-thought architecture and native function calling. It is designed to be the definitive platform for enterprise-grade autonomous agents and complex logical deduction.

Capabilities

Qwen 3.6 introduces several breakthrough features for the agentic era:

  • Always-on Thinking Mode: Native chain-of-thought reasoning that provides deep logical deduction for math, coding, and strategic planning.
  • Preserve Thinking Parameter: A unique feature allowing the model to retain and reuse its reasoning context across multi-turn interactions.
  • Native Function Calling: High-reliability tool-use and API orchestration for autonomous project execution.
  • Global Multilingual Reasoning: Optimized for over 150 languages, with specialized strength in technical and cultural translation.
  • Agentic Benchmarking Leader: Top performance on SWE-bench Pro and Terminal-Bench 2.0, showcasing its ability to operate across digital environments.

Technical Specifications

  • Model Sizes:
    • Qwen3.6-Max: Proprietary flagship with 256K context window.
    • Qwen3.6-Plus: High-performance model with 1M context window.
    • Qwen3.6-Flash: Low-latency model for real-time applications.
  • Context Window: 256,000 tokens (Max) / 1,000,000 tokens (Plus).
  • Architecture: Advanced Mixture-of-Experts (MoE) with native reasoning tokens.
  • API Availability: Alibaba Cloud Model Studio and Qwen Studio.
  • Knowledge Cutoff: January 2026.

Use Cases

Qwen 3.6 is a versatile foundation for:

  • Global Enterprise AI: Deploying localized AI assistants across diverse international markets.
  • Complex Software Engineering: Utilizing Qwen 3.6 Coder for advanced codebase management and autonomous debugging.
  • Deep Scientific Research: Leveraging its reasoning capabilities for data synthesis and hypothesis generation.
  • Real-time Multimodal Interaction: Building apps that can see, hear, and talk to users with minimal latency.

Limitations

  • Latency in "Thinking" Mode: While powerful, the deep reasoning mode can have higher latency compared to the standard mode.
  • Potential for Hallucinations: Like all LLMs, it can occasionally generate plausible but incorrect information, especially in highly specialized domains.

Pricing & Access

Qwen 3.6 is primarily accessible via Alibaba Cloud's enterprise platforms:

API Pricing (per 1M Tokens)

  • Input: ~$1.30
  • Output: ~$7.80

Access Options

  • Alibaba Cloud Model Studio: The primary hub for managed API access.
  • Qwen Studio: Direct interface for testing and deploying Qwen models.
  • Open Source: Smaller versions (7B, 14B, 72B) continue to be available on Hugging Face and GitHub under the Apache 2.0 license.

Ecosystem & Tools

Community & Resources

Frequently Asked Questions

The flagship Qwen3.6-Max-Preview was officially released by Alibaba Cloud on April 20, 2026, as a proprietary, high-intelligence model.
No, while previous Qwen models were open-source, the 3.6-Max series is currently a proprietary, hosted offering available via API.
It is a native chain-of-thought architecture that allows the model to process complex reasoning steps internally before providing a finalized answer.
Qwen3.6-Max supports a 256K token context window, while the Qwen3.6-Plus variant supports up to 1 million tokens.

Related Models

DeepSeek V4

DeepSeek's next-gen model featuring unified hybrid reasoning and a 512K context window, setting new open-weights benchmarks for AGI-level intelligence.

Gemma 4

Google's newest generation of open-weight models as of April 2026, featuring advanced reasoning, native agentic orchestration, and frontier-level.

GPT-5.4

OpenAI's most advanced AI model, featuring superior reasoning (Thinking mode), multi-agent orchestration, and native multimodal capabilities.

Hunyuan 4

Tencent's latest flagship AI model family as of early 2026, featuring superior multilingual reasoning, integrated 3D generation capabilities, and native.

Kimi K2.6

Moonshot AI's next-generation flagship released on April 20, 2026, featuring a 1T parameter MoE architecture and optimized Agent Swarm orchestration for.

Llama 4

Meta's open-source Llama 4 family and the proprietary Muse Spark, leading the open-weights movement and closing the gap with frontier models.

Explore More Models

Discover other AI models and compare their capabilities.

Browse Models← Back to Models