02/05 2026
542
Source 丨 Shenlan Finance
1
First, Let's Debunk a Major Myth
Lately, the robotics community has been abuzz not with who will grace the Spring Festival Gala, but with who won't.
One of the most outrageous claims circulating is:
To secure a solo spot at the Horse Year Spring Festival Gala,
Zhiyuan Robotics supposedly offered 60 million yuan,
while Unitree upped the ante to a staggering 100 million yuan.
Sounds thrilling, right? But this is nothing more than a rumor.
Zhiyuan's official statement clarified two key points: Firstly, they did not engage in any bidding process; secondly, they have explicitly stated they will not participate in the 2026 Horse Year Spring Festival Gala. Citing budget constraints, they prioritize R&D and embodied intelligence.
This stance deserves more serious consideration than the question of "who's on the Gala."
2
What Does a 100 Million Yuan Gala Appearance Really Mean?
Investing 100 million yuan for a single appearance at the Spring Festival Gala equates to the annual salary cost of a top-tier core R&D team of 80 to 100 people for a startup still grappling with engineering challenges.
Even for the listed company UBTECH, its annual R&D expenses, as revealed in public financial reports, are only in the hundreds of millions yuan.
Trading a 4-hour national spectacle for the computational output of 100 top minds over a year is an extremely poor return on investment.
Instead of joining the Gala, Zhiyuan plans to host its own global live broadcast, "The Wonderful Night of Robots." Led by hundreds of robots, the show will feature models that are either fully mass-produced or near-mass-production.
While it may resemble a performance, the underlying logic is vastly different. The Spring Festival Gala is a themed event where safety and festivity take precedence; whereas the self-hosted live broadcast is essentially a technical showcase.
The focus shifts from synchronized dance steps to stable force control, smooth group coordination, and seamless interactions. In essence, it's a serious demonstration of true capabilities.
3
Back to CES: The Anxiety Behind the Prosperity of Homogenization
This brings my attention back to Las Vegas a month ago.
At CES, Chinese humanoid robots were undoubtedly the most talked-about "stars." Unitree G1 was throwing punches, Zhongqing was performing backflips, and their booths were swarmed with visitors.
But from an industry observer's perspective, a sense of emptiness lingers—it seems these robots are desperately trying to prove to the outside world: "I can move, I resemble humans, and I'm affordable."
Conversely, what truly made global engineers wait in line for 40 minutes were Boston Dynamics' Atlas and Figure.
They barely showed off flashy moves; instead, they directly demonstrated their applications in factory scenarios.
4
In-Depth Dialogue: What Lies Ahead?
Against this backdrop, Zhiyuan's choice becomes evident.
It's a typical shift from the virtual to the real.
In fact, similar choices have been made earlier. After Unitree's popularity surge following last year's Spring Festival Gala, a Shenzhen-based robotics company told us they had the opportunity to participate but ultimately declined, opting to allocate resources to engineering implementation instead. Today, their robots are operating in automotive factories.
If the stage is no longer the Spring Festival Gala or the CES booth, where does the next stage lie for Chinese humanoid robots?
With this question in mind, during CES, we specifically spoke with Chen Xinyi, co-founder of Silicon Valley Park, who has long focused on the embodied intelligence track.
Let's begin with her controversial judgment.
In her view, Chinese humanoid robots are firmly in the global first tier. Whether in terms of form, function, design, or overall technological sophistication, China is at the forefront.
So why do we often see "dancing, punching, and backflipping"?
Her explanation is quite measured: These seemingly entertainment-oriented demonstrations are essentially a faster way to capture public attention and serve as a form of "technological popularization"—helping more people understand what robots can and cannot do, conveying a sense of "harmlessness," and reducing the public's natural fear of AI and robots. More importantly, the ability to dance and punch indicates that the technology has reached this stage. Meanwhile, the truly industry-oriented aspects—such as dexterous hands, precise operations, and scenario adaptation—are already quite mature, just not as "visually appealing" or easily shared as short videos.
Looking ahead, her judgment is very clear: She believes the next step for Chinese humanoid robots is to move more smoothly, enhance safety, and truly land in specific scenarios like handling and transportation.
But to achieve this, a practical issue must be addressed—extensive secondary development and functional exploration. This is precisely what Silicon Valley Park is doing: by selling, leasing, and offering secondary development services, they enable more people to access robots at lower costs and collaboratively refine "usability" in real-world scenarios.
We also discussed Boston Dynamics, which many compare themselves to.
In her view, Boston Dynamics' path is not "mysterious" but an inevitable direction.
Industrial robots have long existed; now, they're just becoming lighter, more human-like, and gradually approaching real-life and work scenarios. Whether it's Boston Dynamics or other humanoid robot brands, they will all ultimately converge toward the same goal of "being truly used."
5
Conclusion: A Sobering Judgment
At this juncture, humanoid robots have reached a watershed.
On one side, Optimus is betting on general intelligence; on the other, Chinese manufacturers are actively diversifying after the initial hype.
Some continue to be "stars of the show," while others choose to invest in "brains" rather than "faces."
So here's our judgment: Any "stage appearance" not aimed at real-world scenarios is essentially delaying confronting engineering realities.
For every Spring Festival Gala show skipped, there's an additional round of engineering iteration. This kind of "stinginess" might just mark the true maturity of Chinese humanoid robots.
Shenlan Finance New Media Cluster originated from the Shenlan Finance Journalist Community and boasts a 15-year history. As a renowned domestic financial new media outlet, its accounts focus on China's most valuable companies, cutting-edge industry developments, and emerging regional economies, providing valuable content for investors, corporate executives, and the middle class. Welcome to follow.