9,998 yuan: Humanoid robots are now being sold for under 10,000 yuan.
```
Author | Huang Yu
Editor | Wang Xiaojuan
Over the past three years, the humanoid robot sector has remained hot, and industrial commercialization is accelerating. However, due to insufficient product maturity and high prices, humanoid robots are either research tools in laboratories or performers in high-end commercial scenarios, still far from "entering ordinary households."
Now, this situation is about to see a significant turning point. On October 23, Songyan Power officially launched the pre-sale of Bumi Xiaobumi, the world's first high-performance humanoid robot priced under ten thousand yuan, with a price tag of 9,998 yuan, bringing a "real" humanoid robot into the mass market for the first time.
This marks that humanoid robots are shifting from laboratories and high-end commercial use into the consumer era.
At the media briefing that day, Jiang Zheyuan, founder and chairman of Songyan Power, told Wallstreet CN that the core purpose of releasing the new product was to lower the price to under ten thousand yuan so that more consumers could try it out, to boost sales in the consumer market, and to further reduce supply chain costs.
However, he also admitted that Xiaobumi is still primarily aimed at performance and educational use, and cannot yet meet people's expectations of having robots do real work. He agrees with the prediction of Sergey Levine, a professor at UC Berkeley and top robotics expert: by 2030, robots will be able to independently manage an entire household like a housekeeper.
The "dawn" of large-scale industrialization of humanoid robots is getting closer, and Songyan Power is trying to accelerate the process by lowering Xiaobumi's price. Meanwhile, the first shot in the consumer humanoid robot market has been fired, and competition will only intensify from here on.
How to "Forge" a Thousand-Yuan Machine
Songyan Power has always focused on developing small-sized humanoid robots. Xiaobumi, released this time, is also small—94cm high, and weighs only 12kg.
The low price is Xiaobumi's biggest selling point. In the current humanoid robot industry, most products cost hundreds of thousands of yuan; robots priced at tens of thousands are already considered low-price. Xiaobumi, for the first time, puts the price below ten thousand, making humanoid robots affordable to more people.
Although the price is under ten thousand, Xiaobumi is not a "toy"—Songyan Power calls it "high performance." According to official introduction, Xiaobumi has excellent motion control, can walk stably and smoothly with two legs, and can even perform flexible dance moves.
From a product positioning perspective, Xiaobumi is mainly targeted at technology enthusiasts, teenagers learning to program, and family users, as well as offering kits and course solutions to educational institutions and maker spaces.
In other words, Xiaobumi's main applications remain education and entertainment; it has not yet realized expectations for household chores.
Jiang Zheyuan said humanoid robots have not entered the consumer level because the technology is not there yet, and models are not general enough to handle elderly care or domestic work. "Our idea with Xiaobumi is: if humanoid robots can't do those jobs yet, let's first provide emotional value—accompany children, teach programming, learn English. This is the fastest way into households in the short term."
Therefore, Xiaobumi can be considered a quasi-consumer-grade humanoid robot, standing between consumer and non-consumer realms.
Regarding the choice to develop small-sized humanoid robots, Jiang Zheyuan noted three main reasons: first, to increase the robot's friendliness and reduce user fears; second, to make cost-effectiveness extreme; third, in the physical world, smaller robots have better athletic ability.
Small-sized humanoid robots are not rare within the industry. Besides Songyan Power's previously released N2, E1, etc., there are also UBTECH's Wukong series, Leju's small-sized robots, etc. But Jiang Zheyuan emphasized that, within ten thousand yuan, Xiaobumi's performance is a historical first.
How does Songyan Power balance cost and performance to compress the price of a "high-performance" humanoid robot to below ten thousand?
According to Wallstreet CN, several factors enabled Xiaobumi's under-ten-thousand price: extensive use of composites (instead of all metal), for both light weight and low cost; low product weight reduces costs; high degree of in-house R&D—controllers, entire motor systems are all developed by Songyan, reducing core component costs; and product gross margins are low.
"For us, humanoid robots are essentially plastic, aluminum, iron, copper wire, magnets, and chips—nothing special, so the basic price can be brought down." Jiang Zheyuan revealed that the core components of Xiaobumi are almost 100% domestically produced; the standard version has fully replaced imports, and the second generation will likely be even cheaper.
Jiang Zheyuan believes that, in the future, all humanoid robot makers' prices will go down until stabilized at reasonable profit rates and reasonable price levels. As prices decrease, there will be two main impacts.
On the one hand, more use cases can be explored; on the other hand, industry penetration will accelerate. Lower prices from main brands will drive supply chain costs down, making the industry more accessible. With lower costs for everyone, different forms of robots can be used in different scenarios.
"For example, with Xiaobumi's debut, upstream suppliers for joints, harnesses, etc., also lower costs, robotic arms get cheaper, and small manufacturers can afford them." said Jiang Zheyuan.
Commercialization Reaches an Inflection Point
Even though well-known investor Zhu Xiaohu criticized the sector for "unclear commercialization" and questioned "Who would spend over a hundred thousand to buy a robot for these jobs?", this has not halted the industry's progress towards large-scale commercialization.
2025 is expected to be the first year of mass production of humanoid robots, with companies like Unitree, Zhiyuan, Figure, 1X, and UBTECH making continuous progress. The top OEM brands are shipping more than a thousand units, and the industry is accelerating from laboratory demonstrations to real-world application and batch delivery.
Many large orders for humanoid robots are emerging.
In July this year, Zhiyuan Robotics and Unitree Technology won the bid for China Mobile's (Hangzhou) Information Technology Co., Ltd. (fully owned by China Mobile) procurement project for humanoid biped robot OEM services, with a total budget of 124.05 million yuan (including tax). Among previously disclosed public orders, this is the largest domestic order for humanoid robots.
A Guotai Haitong Securities research report pointed out that this procurement and bidding is an important milestone in the commercialization of China's humanoid robot industry, with the industry currently at the critical "zero to one" stage.
UBTECH, known as the "first stock in humanoid robots," set a global record in September with an order of 250 million yuan. In October, it won another order for the "Guangxi Embodied Intelligence Data Collection and Testing Center Equipment Procurement and Installation" project worth 126 million yuan.
At Tesla's third quarter 2025 earnings call, CEO Elon Musk revealed that the Optimus humanoid robot is planned to show a near-production prototype in Q1 2026, with a target of one million units annual capacity in the future.
Although it has been established for just over two years, Songyan Power has also made good commercial progress.
Songyan Power disclosed that in the first half of this year, it received over 2,000 commercial orders, with total contract value expected to exceed 100 million yuan. This means Songyan Power has become, after Unitree, the second domestic company to clearly cross the "thousand units sold" threshold for humanoid robots.
Songyan Power has a clear commercial strategy. Jiang Zheyuan said: "For the second half of this year, our biggest goal is to do a good job in production and delivery, and turn the over-100-million-yuan in orders into confirmed revenue. Next year, our goal is to deliver ten thousand units."
Humanoid robots are achieving the leap from "technically feasible" to "commercially feasible," but there is still some distance to real entry into thousands of homes.
Zhang Zhengyou, Tencent's chief scientist and Director of Robotics X Lab, believes commercialization of humanoid robots is far from the "cell phone" era; most are still used for data collection, research, or as guides.
Some optimists believe the humanoid robot boom will come before 2030, while others see it coming after. UBS Securities China industrial analyst Wang Feili said a key milestone will be the "electric car moment"—when technical bottlenecks are broken and sales leap from one million to ten million units within five years. For humanoid robots, such a moment is "unlikely in the next five years."
For humanoid robots to truly do work, the industry widely sees the main blocker as insufficient embodied intelligence data. Jiang Zheyuan pointed out that the scale of data is currently inadequate, and many governments and listed companies are building embodied intelligence training grounds to collect data, reflecting a serious "data drought," with little improvement this year.
Yao Maoqing, partner, senior VP, and president of the embodied business unit at Zhiyuan Robotics, also said that what matters most for Zhiyuan now is letting robots enter as many real scenarios as possible, to obtain high-quality data—this is key for their current technical approach.
Since the start of the year, the embodied intelligence sector—including humanoid robots—has continued to boom, attracting continuous capital and industrial enthusiasm. Thanks to events like marathons, boxing, and soccer, humanoid robots are rapidly refreshing the public's perception.
It can be expected that as more companies launch competitive humanoid robots, competition in this sector may reach a fever pitch next year. Only those with leading technology, capital support and commercialization ability will survive rounds of tests and become potential winners.
Risk Warning and DisclaimerThe market has risks, investment needs caution. This article does not constitute personal investment advice, nor does it consider the special investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, viewpoints or conclusions herein are suitable for their specific circumstances. Investment based on this article is at your own risk. ```