Alibaba cuts Qwen AI model costs on Qoder coding platform to capture US workday demand
Alibaba Group Holding has slashed prices for its flagship Qwen AI models by up to 80 per cent on its Qoder agentic coding platform, offering an aggressive discount to hook global developers to compete with US and Chinese rivals like Anthropic and Zhipu AI. The e-commerce and cloud giant said on X on
By Xinmei Shen

Alibaba Group Holding has slashed prices for its flagship Qwen AI models by up to 80 per cent on its Qoder agentic coding platform, offering an aggressive discount to hook global developers to compete with US and Chinese rivals like Anthropic and Zhipu AI.
The e-commerce and cloud giant said on X on Tuesday that it was cutting the price of its flagship Qwen3.7-Max model by 80 per cent and its smaller Qwen3.7-Plus model by 60 per cent for international users from 10pm to 8am Beijing time. The promotion is ongoing and will run during the same time daily.
While the promotion is branded as an “off-peak” discount, the timing – 10am to 8pm US Eastern Time – appears to be a calculated move to capture American users.
“If you’re in the Americas, here’s the twist: off-peak covers most of your workday,” Alibaba said in a post on X, the social media site formerly known as Twitter.
In China, the discount applies to the platform’s Qoderwork tool, which supports AI models including Qwen.
To capitalise on the discount, the company suggested Chinese users schedule tasks during the day, or queue commands before bed, allowing Qoder’s autonomous agents to execute code overnight.
The promotion follows a separate half-price discount for Qwen3.7-Max that concluded earlier this week.
The aggressive pricing comes amid intense rivalry between US and Chinese artificial intelligence labs to capture users looking to automate complex coding workloads.
Chinese labs are also looking to capitalise on US regulatory restrictions.
When US-based Anthropic recently blocked global access to its cutting-edge Claude Fable 5 model following a Washington directive, Chinese rival Zhipu AI stepped into the vacuum by releasing GLM-5.2 – an open-weight model close to US coding capabilities, according to several benchmarks – with founder Tang Jie marketing it as “frontier intelligence” that “belongs to everyone.”
Alibaba is positioning its closed-source Qwen3.7-Max as a heavy-duty alternative. Launched last month, the model “excels at agentic coding, complex reasoning and long-horizon tasks”, according to Alibaba. It also said that the model could autonomously operate for up to 35 hours without performance degradation.
Alibaba is gearing up to be China’s “AI factory” that operates “all five layers of the full AI stack”, Liu Weiguang, senior vice-president of Alibaba’s cloud computing unit, said last month. The five AI layers are chips, agentic cloud, AI models, model service platforms and agentic applications, he said.
Alibaba chairman Joe Tsai said last week at the VivaTech conference in Paris that AI could ultimately represent a US$50 trillion market and vowed to invest across the industry’s entire value chain.
Alibaba owns the South China Morning Post.
