Китайские ИИ-стартапы остаются позади: высокий риск и низкая прибыль Translation: Chinese AI Startups Lag Behind: High Risk and Low Profit

Chinese startups are significantly trailing behind their American counterparts in terms of revenue. This is highlighted in a report by Unique Research and the San Francisco-based consulting firm Tech Buzz China, as reported by SCMP.

As of August, only four AI applications from private companies in China made it into the top 100 for annual recurring revenue. Collectively, Glority, Plaud, ByteDance, and Zuoyebang generated $447 million, merely 1.23% of the total revenue of $36.4 billion.

Notably, this research group did not include applications from publicly traded giants like Alibaba Group Holding and Tencent Holdings.

Rui Ma, co-founder of Tech Buzz China, attributed the disparity to the structure of China’s economy, where local startups prioritize quick profits, often favoring short-term projects such as those in the public sector (government contracts and municipal bids).

«Attracting global consumers and businesses demands significantly more capital,» she mentioned.

Ma also noted that Chinese companies find it more challenging to raise funds in domestic markets.

The most profitable among Chinese firms is Glority from Hangzhou, known for its popular plant identification app, PictureThis, which achieved an annual revenue of $173 million, ranking 20th overall.

Leading the list are American startups OpenAI and Anthropic, with revenues of $17 billion and $7 billion, respectively.

Delta Wu, head of Unique Research, pointed out that the evaluations are based on internal models that take into account traffic redirected from each product’s official website to third-party payment platforms like Stripe.

Nineteen out of the top 23 most profitable Chinese apps derive a substantial portion of their revenue from overseas, indicating potential growth in those markets, according to Ma.

While American companies report higher revenues compared to their Chinese competitors, they are also making multi-billion-dollar investments to develop data centers.

Morgan Stanley analysts predict that over the next five years, the total capacity of data centers will increase sixfold. They estimate that by the end of 2028, the combined value of data centers and related equipment will reach $3 trillion.

An April report from McKinsey indicated that meeting the rising demand for AI infrastructure will require capital investment of approximately $5.2 trillion by 2030. In contrast, spending on data centers for traditional IT applications is estimated at $1.5 trillion.

Current revenues from AI firms may not justify the enormous costs associated with computing, according to HSBC CEO Georges Elhedery.

Another challenge is the high energy requirement of data centers. Rising demand is leading to increased tariffs, which is causing dissatisfaction among regular consumers.

Remember, in October, analysts named energy the most valuable resource in the world due to the AI boom.