Alibaba Qwen3.8-Max: 2.4T Params, Powers AI Dev

Alibaba has launched Qwen3.8-Max, a new multimodal artificial intelligence model. This AI can handle coding, visual analysis, and long-running tasks. What's interesting is its design. It has 2.4 trillion total parameters but only uses 95 billion for each operation. This helps reduce computing needs and latency. The model supports a context window of up to one million tokens, letting it process large amounts of text and visual material. It can even turn long documents or livestreams into searchable knowledge bases. Alibaba states Qwen3.8-Max ranked highly in various benchmarks, including Text Arena and Vision Arena. It's available to developers through APIs on Alibaba Cloud Model Studio and QwenWork, the company’s workplace AI agent platform. The company plans to release the model weights next week. Alibaba is positioning Qwen3.8-Max for autonomous software development. An internal test showed it worked for 16 days on a software engineering project, leading to the open-source release of "oh-my-cli." This new model also outperformed human participants in a multimodal dialogue challenge and can generate front-end projects from screenshots, create 3D models from floor plans, and edit video footage. The bottom line is this model aims to advance AI capabilities for complex and extended tasks, potentially changing how software is developed and how we interact with AI for creative and analytical work.