At the Chuangzhi Cloud City in Shenzhen, the first Xin Summit showcased a variety of unexpected smart products—from tofu machines to barbecue grills, ice makers to cocktail shakers. Chen Yu, a partner at Yunqi Capital, stated during a roundtable forum: “Almost every device you can think of is being made smart and AI-enabled.”
By 2025, investment data in AI hardware began to recover, with over fifty percent of funding in May alone flowing into the AI hardware sector. However, in this era where “everything can be AI,” technology investors are not feeling exuberant but rather anxious.
“We hustle every day and feel very anxious,” admitted Liang Yu, founding partner of Chuangshi Partners (CCV). He pointed out that market competition has become fierce, with concerns that new models may disrupt the existing landscape, as well as worries that entrepreneurs’ directions are too fragmented to scale in the future, and even fears that the entire sector may be eliminated by new technologies.
At the XIN Summit, a student entrepreneurial team showcased an AI children’s aquarium, indicating that smart products are entering various niche markets.
The Hot Market for AI Hardware in China
In the global wave of AI entrepreneurship, the financing heat for consumer-grade AI hardware is particularly significant in the Chinese market.
This is attributed to the national support for embodied intelligent AI and a strong supply chain. Chen Yu emphasized that in Shenzhen, one can find all the components needed for a product within two hours, whereas in the United States, it may take weeks or even longer.
In 2025, Yunqi Capital invested in various consumer scenario AI hardware projects, including pool cleaning, smart imaging, and smart translation, focusing on the practical application value of these technologies in daily life.
Chen Yu pointed out that this wave of AI hardware entrepreneurship exhibits different characteristics from previous ones. Firstly, the variety of entrepreneurial categories is extremely rich, and the barriers to entry are relatively low, with many unexpected smart products visible in Shenzhen.
Plaud Note is a successful example of a company that has shipped nearly 700,000 units since its establishment in 2021 without any external financing.
This “fragmented” and “cash-rich” market characteristic has prompted investment institutions to change their strategies. Chen Yu stated that investors need to conduct a “carpet-style” investigation like salespeople, analyzing potential hardware companies one by one to find valuable targets among numerous niche categories.
Solving Real Problems is Key
Investors believe that the logic of AI hardware products in the current market is already very clear: one type uses AI as an auxiliary function, while the other treats it as a core necessity.
Liang Yu pointed out that the product logic from the past internet era still applies in the AI era, which is to solve core problems that users are willing to pay for and create stickiness.
Many investors agree on one point: the success of AI hardware must return to first principles, namely whether it truly solves real problems and provides an exceptional technical experience. Otherwise, the product either does not solve a real problem, or the existing solutions are already good enough.
Xu Chen, managing partner at Gobi Partners, stated that he would not make heavy bets before seeing a clear endgame. He believes the most reliable approach is to look for “AI+” opportunities in traditional industries, truly understand customer needs, and actively promote business.
He also emphasized that AI hardware based on large models still faces challenges such as cost fluctuations, with larger single financing scales that could create bigger “bubbles” if problems arise.
Another survival path is to go overseas. Liang Yu cited AI recording pens as an example, noting that such products typically gain traction in the U.S. first, then quickly form a second major market in Japan, ultimately achieving globalization.
This logic has been validated in reality. Plaud.AI achieved tens of millions of dollars in revenue in its first battle in the U.S. market, realizing self-sufficiency through a “hardware sales + software subscription” model.
Liang Yu predicts that in the next two years, there will be a large number of mergers and integrations in the industry. For Chinese entrepreneurs, they should actively target the U.S. market, leveraging the supply chain capabilities of the Greater Bay Area to “earn dollars” in software subscriptions, new retail, and other fields.
This may be the most worthwhile solution for Chinese AI hardware entrepreneurs to explore in 2025, amidst anxiety and internal competition.