Alibaba Unveils Qwen3.8-Max as Open-Source AI Model With 2.4 Trillion Parameters
Table of Contents
Alibaba has introduced Qwen3.8-Max, its latest flagship artificial intelligence model featuring 2.4 trillion parameters. Announced on Monday, the model marks the company’s return to an open-source strategy after several recent flagship AI releases remained proprietary. Following the announcement, Alibaba’s Hong Kong-listed shares rose sharply in early trading.
Alibaba Returns to Open Source
Alibaba said it plans to release the model’s weights for public download next week, making Qwen3.8-Max the first Max-series model in the Qwen family to become open source. The company aims to give developers and enterprises broader access to its latest AI technology while continuing to expand its cloud-based AI services.
Multimodal AI With Enterprise Features
According to Alibaba, Qwen3.8-Max supports a context window of up to one million tokens and activates 95 billion parameters from its total of 2.4 trillion parameters during inference.
The multimodal model can process long documents, videos, and live streams to create searchable knowledge bases. Alibaba also said the model can recreate software applications from screenshots, generate interactive games and educational animations, convert two-dimensional floor plans into 3D visualisations, and support autonomous coding, complex research, and other long-running agentic tasks.
Benchmark Performance
Alibaba said benchmark testing showed Qwen3.8-Max was broadly competitive with leading models from OpenAI and Anthropic. According to the company, the model outperformed competing systems on several coding, multimodal, and engineering benchmarks while trailing on some general-purpose reasoning evaluations.
Bloomberg also reported that Qwen3.8-Max ranked higher than Moonshot AI’s recently introduced Kimi K3 on some benchmark tests.
Competition in China’s AI Market
The launch comes as competition among Chinese AI developers continues to accelerate. Companies including DeepSeek, Moonshot, and ByteDance have introduced new AI models in recent months, increasing pressure across the industry.
Earlier this year, DeepSeek’s low-cost reasoning models reshaped the competitive landscape and encouraged rival companies to release updated AI systems more rapidly.
Expanding AI Investment
Alibaba has continued increasing its investment in artificial intelligence to strengthen its position in the market. In May, the company announced that it would exceed its previously planned investment of 380 billion yuan (approximately $55.96 billion) in AI infrastructure over three years.
The company said Qwen3.8-Max will be available globally through Alibaba Cloud’s Model Studio APIs and QwenWork, its workplace AI agent platform for enterprise users.
Key Highlights
| Feature | Details |
| Model | Qwen3.8-Max |
| Parameters | 2.4 trillion (95 billion active) |
| Context Window | Up to 1 million tokens |
| Model Type | Multimodal |
| Open Source | Model weights planned for release next week |
| Availability | Alibaba Cloud Model Studio APIs and QwenWork |
| Main Uses | Coding, research, multimodal tasks, knowledge bases |
Conclusion
The launch of Qwen3.8-Max highlights Alibaba’s renewed focus on open-source AI while expanding its enterprise AI capabilities. With support for one million tokens, multimodal processing, and advanced coding and research features, the model is designed for demanding AI workloads. As competition among Chinese AI developers continues to intensify, Alibaba’s latest release reinforces its commitment to both open innovation and commercial AI services.
