Alibaba Launches Qwen 3.8-Max, Its Largest AI Model Competing With US Leaders

By: Harsh Vardhan

|

Updated on: 03-Aug-2026 06:00 PM

7,556 views

Follow Us:

7,556 views

Alibaba has introduced Qwen 3.8-Max, its largest and most advanced AI model to date. The company claims this new model matches the performance of leading US AI systems, including OpenAI's ChatGPT and Anthropic's Claude, in coding, reasoning, and multimodal tasks. Alibaba also offers Qwen 3.8-Max at lower prices for developers.

Key Highlights

• Alibaba unveils Qwen 3.8-Max, its largest AI model with 2.4 trillion parameters.
• Model supports up to one million token context window and advanced multimodal tasks.
• Qwen 3.8-Max priced at US$2 per million input tokens and US$6 per million output tokens.
• Model can handle autonomous coding, long-term planning, and complex reasoning tasks.

Qwen 3.8-Max Features and Capabilities

Qwen 3.8-Max is built on 2.4 trillion parameters and supports a context window of up to one million tokens. This allows the model to process lengthy documents, large code repositories, images, and videos in a single prompt. Developers can access Qwen 3.8-Max through Alibaba Cloud's Model Studio APIs, while enterprise users can use it via QwenWork, Alibaba's AI workplace assistant.

Alibaba highlights the model's autonomous capabilities. Qwen 3.8-Max can independently manage software engineering projects, perform long-term planning, and complete coding tasks with minimal human input. In an internal demonstration, the model completed a software engineering project over 16 days, handling planning, coding, debugging, and optimization.

Beyond coding, Qwen 3.8-Max can analyze documents, images, and videos to build searchable knowledge bases. It can recreate software applications from screenshots, generate interactive educational content, and convert two-dimensional floor plans into 3D visualizations.

Performance, Pricing, and Market Context

Alibaba states that Qwen 3.8-Max does not use all 2.4 trillion parameters for every task. Instead, it activates around 95 billion parameters as needed, which helps deliver faster responses and reduces computing costs.

The company has priced Qwen 3.8-Max at US$2 per one million input tokens and US$6 per one million output tokens. Implicit caching costs US$0.25 per one million tokens. These rates make Qwen 3.8-Max one of the more competitively priced frontier AI models currently available.

According to Alibaba, Qwen 3.8-Max performs as well as, or better than, leading models from OpenAI, Anthropic, and Google across various industry benchmarks. The model reportedly excels in coding, software engineering, research, long-video understanding, and computer-use tasks. Qwen 3.8-Max is also said to be the highest-ranked Chinese text model on Arena.AI and ranks second globally for visual understanding, behind Anthropic's Claude Fable 5. However, these results are based on Alibaba's own data and have not been independently verified.

The launch of Qwen 3.8-Max follows a series of major AI releases from Chinese companies in recent months. Firms such as Moonshot AI, DeepSeek, ByteDance, MiniMax, and Z.ai have all introduced new models as they compete with US AI leaders. Alibaba plans to release the open weights for Qwen 3.8-Max and a smaller Qwen 3.8-27B version next week, allowing developers to download and fine-tune both models.