Alibaba's Qwen3.8-27B: Local AI & Revenue Share

Alibaba is set to release its new large language model, "Qwen3.8-27B," at midnight on August 15th. This AI is designed for local deployment, meaning it can run on consumer hardware like high-end graphics cards and PCs with Ryzen AI Max processors. What's interesting is that this model inherits technology from its predecessor, Qwen3.6-27B, while incorporating performance from the higher-tier Qwen3.8 Max. It offers native multimodal and multilingual support, including image and video understanding. Users can expect capabilities like handling STEM diagrams and complex documents. The company is also exploring a new business model for its open AI. For large commercial users of Qwen3.8-Max, Alibaba plans to require a portion of their revenue. This revenue-sharing mechanism aims to help AI developers recover significant research and development costs. This could be a significant shift in how open-source AI models are monetized.