# Alibaba’s 75% Capex Surge: Driving High-Performance Edge AI Despite Profit Drops > Source: > Published: 2026-08-21 09:00:00+00:00 Alibaba has dramatically increased its capital expenditure by 75% to build out AI infrastructure despite a 76% drop in quarterly net income, while simultaneously releasing highly optimized open-source AsiaAI Publisher · August 21, 2026 · 2 min read AI & Machine Learning Alibaba has dramatically increased its capital expenditure by 75% to build out AI infrastructure despite a 76% drop in quarterly net income, while simultaneously releasing highly optimized open-source Qwen3-8B and Qwen3-27B models. These newly released models achieve frontier-class performance on standard consumer hardware through advanced quantization, demonstrating how Alibaba’s heavy infrastructure spending is directly translating into highly accessible edge-AI capabilities. Together, these developments show the e-commerce giant’s determination to lead both the cloud infrastructure and local-execution tiers of the AI race. For Western observers, Alibaba’s dual strategy of heavy infrastructure spending and aggressive open-sourcing reveals that Chinese tech giants are not being slowed down by US export controls. Instead, they are successfully engineering around hardware constraints by pairing massive domestic capital investments with extreme software-level optimization. For Western observers, Alibaba’s dual strategy of heavy infrastructure spending and aggressive open-sourcing reveals that Chinese tech giants are not being slowed down by US export controls. Instead, they are successfully engineering around hardware constraints by pairing massive domestic capital investments with extreme software-level optimization. Source Qwen3.8-27B Is the New Model Kill Line: Local ‘Opus 4.6’ on Consumer Hardware