Qwen3-Max, the first trillion-parameter model from Alibaba Cloud, was one of the company’s most expensive models on its release in September, with tiered pricing that charged a minimum of US$0.861 per million input tokens and US$3.441 per million output tokens for application programming interface (API) users.
Alibaba Cloud is the AI and cloud services unit of Hangzhou-based Alibaba, owner of the South China Morning Post.
On Thursday, the company reduced prices by almost 50 per cent across the different tiers of usage for the model on its official cloud services platform for domestic users, with the minimum price falling to US$0.459 per million input tokens and US$1.836 per million output tokens.
A further discount of 50 per cent was offered for batch API calls, where tasks are executed during non-peak hours.
Qwen-3 Max recently topped a high-profile cryptocurrency investment competition among six leading US and Chinese models with the highest returns.
