In its November 16 report, the Citi team led by Alicia Yap cut the average cost per task for Alibaba’s Qwen4-Max from RMB 26.58 to RMB 11.16, a 58% drop. Token consumption in benchmark tests fell by 24% to 49%. The model has 2.4 trillion parameters in total, with 71 billion activated, and saves these costs through architectural optimizations such as sparse attention. The report also maintained an Alibaba Buy rating unrelated to $BNB and a $193 price target, implying 109.3% upside from the share price of $92.21. My own approach is to first see whether the price and volume trend can continue; there’s no need to rush into chasing a single message.
