371
Productivity & Workflow357
Software Development253
Automation & Workflow230
Marketing & Growth213
AI Infrastructure & MLOps197
Photography & Imaging170
Writing & Content Creation197
Design & Creative175
Data & Analytics162
Customer Support161
Education & Learning149
Sales & Outreach132
Voice & Speech131
Operations & Admin118
New estimates show Chinese AI companies have far less money and revenue than top US firms, even as their models stay competitive and cheaper.
In short: Chinese AI companies are building capable and often cheaper AI models, but they are falling further behind US leaders in money raised and money earned.
Reports cited by The New York Times point to a widening financial gap between US and Chinese AI companies. One example is Anthropic, a US AI company, which is reported to be pursuing an IPO (a stock market listing) valued around $81.3 billion, plus another $26 billion raised privately.
Venture funding, which is money investors put into young companies, shows a similar split. One estimate says US AI firms raised about $380 billion from 2023 to 2026, while Chinese AI startups raised roughly $39 billion over the same period.
Revenue, which is the money a company brings in from customers, also looks very different. Estimates put OpenAI and Anthropic’s combined annualized revenue at about $105 billion. Another estimate puts the combined annualized revenue of six leading Chinese AI companies at about $11 billion.
At the same time, the cost of building and running AI keeps climbing. Rhodium Group estimates China’s leading tech companies will spend $139 billion on major investments this year, and that could rise above $193 billion in 2027. AI divisions currently account for less than 8% of that spending.
Chinese firms also face tougher fundraising conditions at home, and US restrictions on advanced chips. Chips are the specialized computer parts used to train and run AI, like the engine in a car.
Many Chinese companies are trying to compete with open-weight models (models whose underlying settings can be shared, like a recipe you can copy) and lower prices. The key question is whether they can turn lots of users into steady, repeat income that is large enough to pay for the huge computing costs needed to keep improving their AI.
Source: NYTimes