Chinese AI Giants Struggle to Monetize Advanced Models

China's top technology companies are releasing artificial intelligence models that rival the world's best systems, yet none have solved the fundamental challenge of making money from them.

Companies across China's AI landscape — from established giants like Alibaba and ByteDance to well-funded startups including DeepSeek and Moonshot — are experimenting with different business models to cover the massive costs of building and operating advanced AI systems. The results so far have been mixed, according to reporting from The New York Times.

Alibaba has begun charging users for access to its most capable models, while ByteDance, TikTok's parent company, introduced tiered pricing designed to generate higher revenue from its most advanced offerings. Meanwhile, startups like DeepSeek and Moonshot have raised billions in venture capital to sustain operations.

The economics of AI development

The financial pressure stems from the enormous infrastructure requirements of modern AI. Companies must continuously purchase powerful computing chips — first to build models, then to test improvements, and finally to ensure global performance at scale. They also need to construct or lease data center space to house this computing power.

These economics challenge AI companies worldwide. Silicon Valley leaders including OpenAI, Anthropic, and Google are all investing more in AI development than they currently earn from it.

China's open-source paradox

Chinese companies face an additional complication rooted in their development approach. Most of China's leading AI systems use open-source or open-weight architectures, meaning their underlying technology is publicly accessible.

This openness has accelerated progress across China's AI sector, as companies benefit when competitors share their work publicly. However, it also makes monetization more difficult, as the technology itself becomes harder to protect as a competitive advantage.

Why it matters

The monetization struggle reveals a critical tension in AI development: technical capability doesn't automatically translate to viable business models. For China's AI sector, this challenge could determine which companies survive the current investment cycle and whether the country's open-source approach proves sustainable long-term. The outcome will shape not just China's AI landscape but global competition in the technology.

These details were first reported by Meaghan Tobin for The New York Times.