Alibaba releases Qwen4-Max, says average cost per task falls 60%
阿里Qwen4正式发布:性能升级,性价比更优,单任务平均费用大降6成
Price cuts and model releases feed the Chinese AI competition narrative around Alibaba and its peers.
Alibaba released Qwen4-Max on Nov 16, a 2.4-trillion-parameter base model with native vision and a one-million-token context window, according to Wallstreetcn. API pricing for cache hits starts at 0.5 yuan per million tokens and the average cost of completing a task is about 60% lower than the previous model; smaller Qwen4 models are reportedly to be open-sourced.