Chinese AI firms Z.ai and MiniMax could remain loss-making until 2030, Macquarie says
Chinese AI firms Z.ai (Zhipu AI) and MiniMax are projected to remain unprofitable through 2030, despite anticipated rapid revenue growth. This outlook, from Macquarie Group's head of Asia internet and software research Ellie Jiang, underscores the substantial costs associated with developing and operating frontier AI models. A primary factor is the high expense of computing power, exacerbated in China by US restrictions on advanced Nvidia processors, creating a compute shortage two to three times more severe than the global average. Macquarie's conservative forecast contrasts with Z.ai's internal projection of US$2.4 billion in annual recurring revenue (ARR) by year-end, with Macquarie estimating closer to US$3 billion.
The continued unprofitability of Chinese AI firms Z.ai and MiniMax until 2030, as forecast by Macquarie Group, reflects the immense capital expenditure required for frontier AI development. This is not merely a cost of innovation but a direct consequence of China's acute compute crunch, which is two to three times more severe than the global average due to US export controls on Nvidia's advanced chips. The challenge for Chinese AI developers is clear: how to scale revenue rapidly enough to offset these escalating infrastructure costs. While Macquarie projects significant annual recurring revenue (ARR) growth, with Z.ai's ARR potentially reaching US$3 billion by year-end, the path to profitability remains distant. This situation points to a critical bottleneck for China's AI ambitions, where even high revenue growth may not translate into immediate financial returns. The long investment horizon required for these firms suggests that government support or sustained private capital will be essential to maintain their competitive edge in a technologically constrained environment.
Related reading
6 stories
China targets fourfold boost in AI computing capacity by 2030 in major tech push
China's ambitious plan to quadruple AI computing capacity by 2030 directly addresses the compute crunch highlighted here.
DeepSeek embarks on ‘unprecedented’ hiring spree as it overhauls backend systems
DeepSeek's hiring spree and backend overhaul show another Chinese AI firm grappling with scaling infrastructure and talent.
China’s rocket boom turns Hainan into a space hub. Can launches fuel wider growth?
Anthropic merges Claude chat and Cowork in one interface
Threads’ new features let podcasters promote shows and reach listeners
