On July 2nd, the China Securities Regulatory Commission (CSRC) approved the registration for the IPO of Unitree Technology on the STAR Market.

This approval, coming just 104 days after the application was accepted on March 20th, significantly outpaces the average 150-day review cycle for hard-tech firms on the STAR Market. The 73-day period from acceptance to the listing committee meeting sets a new speed record under the exchange's pre-review mechanism.

This "lightning-fast" approval for an embodied AI company signals one key development: the industry has produced its first verifiable commercial success story.

A Profitable Benchmark

In recent years, the humanoid robot sector has been saturated with concept presentations and product launches. The staggering amounts of venture capital raised are well-documented, yet few companies have delivered solid financial performance.

Unitree presents a markedly different picture. In 2025, the company shipped over 5,500 humanoid robots, ranking first globally. Revenue from its humanoid business reached 868 million yuan, surpassing its quadruped robot segment for the first time to become the primary income source. Total annual revenue hit 1.699 billion yuan, with adjusted net profit reaching 591 million yuan. The gross margin for its core business stood at 60.13%.

Over three years, revenue surged from 159 million to 1.7 billion yuan. The company transitioned from losses to profitability, and its business shifted from being dominated by quadruped robots to having humanoids account for over half of revenue. The sale of 5,500 humanoid units signifies that the industry has finally established a verifiable formula: product definition → mass production and delivery → revenue growth → profitability.

Once this commercial loop is proven, the industry's logic shifts from hypothesis to established fact. After years of conceptual hype, China's embodied AI sector finally has a player that has progressed from technological exploration to a complete business model. The first company to pass the test has set a high bar, which should raise pressure on its competitors.

A Clear Signal

Following the approval, Unitree updated its prospectus. Notably, the top-listed risk factor changed from "failure to achieve technological breakthroughs as expected" to "risks of slowing growth and operational performance volatility."

This change in priority is telling. For years, industry competition centered on which company could make a robot move first—comparing gait, degrees of freedom, and human-likeness. Unitree's successful listing indicates the sector is moving from technical validation to commercial validation. The test has changed; those who cannot adapt will be forced to exit. The new competitive focus will be on sustainable delivery, cost control, and identifying genuine application scenarios.

Unitree itself is pivoting. Of its planned 4.2 billion yuan fundraising, 2.022 billion yuan (48%) is earmarked for embodied large model R&D, making it the single largest investment, far exceeding the 1.11 billion yuan for hardware development. Transitioning from "in-house hardware R&D" to full-stack competition in "brain and body," the company recognizes that hardware sales alone cannot support its 42 billion yuan valuation.

Challenges, however, are evident. In Q1 2026, Unitree's revenue was 423 million yuan, a year-on-year increase of 68%, but its adjusted net profit fell 52% to 40.25 million yuan from 84.83 million yuan. While doubled R&D spending and surging marketing costs are surface-level explanations, the deeper reasons are the natural slowdown in growth from a larger revenue base and increased competition—with Tesla's Optimus Gen-3 in small-scale production and Xiaomi's CyberOne planning deployment of 2,000 units in automotive plants.

Unitree acknowledges in its prospectus that its product assembly relies heavily on manual labor. In contrast, Tesla and Xiaomi possess vast automotive factories as training grounds. Unitree's advantage lies in its earlier development of the robotic body, while its disadvantage is the lack of sufficient real-world scenarios for professional embodied AI training.

A Valuation Benchmark

The significance of Unitree's listing for the capital market lies first in establishing a valuation anchor.

Based on the 4.2 billion yuan fundraising target and an estimated 10% issuance ratio, the implied market capitalization at listing is approximately 42 billion yuan. Founder Wang Xingxing, holding about a 33% stake, would see his personal wealth reach around 14 billion yuan. Starting a decade ago with 100,000 yuan, this entrepreneur born in the 1990s has built a workshop into the world's leading shipper of humanoid robots.

The entire sector now has a price reference. Companies valued below this benchmark have a new comparison point, while those valued above it must demonstrate they are stronger than Unitree.

More noteworthy is the regulatory stance. While the average STAR Market review takes about 150 days, Unitree's process took only 73. This year's Lujiazui Forum explicitly included embodied AI and general AI within the scope of the STAR Market's fifth set of listing criteria. Completing the process in 104 days signals regulators' view that embodied AI is a key focus for "new quality productive forces."

However, institutional analysts caution that Unitree's listing will accelerate sector differentiation, with capital increasingly concentrating on "hard assets" that have confirmed orders and mass-production capabilities. Companies reliant purely on thematic speculation will face growing difficulties.

Concluding Thoughts

The IPO approval for Unitree signifies that the embodied AI industry has officially crossed a major threshold.

The era where a simple walking video could secure funding is over. The capital market will now focus on shipment volumes, gross margins, real-world application, and customer retention—not just demos. Unitree's peers, such as Deep Robotics and Leju Robotics, are also advancing their IPO plans. UBTech recently launched its U1 model, and Zhiyuan Robotics is pushing towards a delivery target of ten thousand units.

While the sector has produced its first listed company on the A-share market, the real competition is just beginning. The battle used to be about who could tell the better story; from now on, it will be about who can generate profits and survive under the new rules.