Unitree Robotics has become one of the first humanoid robot makers to go public while actually turning a profit, clearing a $618 million initial public offering on Shanghai's STAR Market. The approval, granted on July 3, 2026, marks a watershed moment in an industry that has long been defined by venture-backed startups burning cash in pursuit of the next breakthrough. Unlike most of its competitors, Unitree reported 2025 revenue of approximately 1.69 billion yuan (about $250 million) and adjusted profit of roughly 591 million yuan (about $87 million), according to filings reviewed by multiple outlets.

The timing of Unitree's IPO is striking because it arrives amid an unprecedented wave of humanoid funding across Asia and North America. Within a two-week window in mid-July 2026, two other major humanoid startups announced nine-figure raises: LimX Dynamics closed a $200 million pre-IPO round at a $2.2 billion valuation, while Walden Robotics, a spinout from Toyota Research Institute, emerged from stealth with $300 million in funding at a $1.1 billion valuation. This convergence of capital suggests that investors believe humanoid robots are moving from research projects into production work, even as the broader market remains skeptical about when these machines will truly scale.

Why Does Unitree's Profitability Matter So Much?

In the humanoid robotics space, profitability is almost unheard of. Unitree's competitor UBTech Robotics, which already trades on the Hong Kong stock exchange at a market cap of roughly 54.8 billion Hong Kong dollars, reported similar 2025 revenue of about 2 billion yuan but posted a 700 million yuan net loss. This comparison reveals a fundamental divide in the industry: some companies have figured out how to build and sell humanoids at a margin, while others are still in a growth-at-all-costs phase. Unitree's ability to achieve profitability while maintaining aggressive R&D spending suggests its business model is genuinely working, not just sustained by investor enthusiasm.

The company plans to sell at least 40.4 million shares, representing a minimum 10 percent stake, to raise approximately 4.2 billion yuan. This implies a valuation of around 42 billion yuan, or roughly $5.9 billion. According to reporting from Caixin Global, the regulatory review process moved at record speed, suggesting Chinese authorities view the humanoid sector as strategically important.

What Are the Key Differences Between Leading Humanoid Makers?

The humanoid robotics market is fragmenting into distinct categories based on what each company has actually achieved. Understanding these differences is crucial for investors and industry observers trying to assess which companies will ultimately dominate:

• Shipment Volume: Unitree has delivered more than 5,500 complete humanoids, making it the leader in pure humanoid production numbers, though much of this volume still serves research and education rather than industrial work.
• Audited Revenue: UBTech has converted full-size humanoids into audited revenue faster than competitors, though its robots are not yet as profitable as Unitree's and the company reports that current robot productivity remains below that of human workers in selected tasks.
• Paid Industrial Deployment: Agility Robotics leads in evidence of repeated, paid work inside real customer facilities, with its Digit humanoid accumulating more than 65,000 operating hours across nine customer sites in manufacturing and logistics.
• General-Purpose Autonomy: Figure AI appears strongest in integrated learned autonomy, particularly for tasks requiring locomotion, manipulation and visual reasoning within a single workflow, though its paid external scale still trails its funding and valuation.

This fragmentation matters because it shows there is no single leader across the entire market. Instead, different companies are winning in different dimensions. Unitree dominates affordable hardware and shipment volume. UBTech leads in audited revenue. Agility Robotics has the strongest evidence of robots performing actual production work. And Figure AI possesses the technology most likely to change the competitive ranking if it can scale paid deployments.

How Are Chinese Manufacturers Gaining an Edge?

China already holds a decisive manufacturing advantage in humanoid robotics. According to estimates cited in the sources, Chinese manufacturers produced roughly 85 percent of the approximately 13,000 global humanoid shipments during 2025. Unitree, AgiBot and UBTech have all crossed meaningful production thresholds, while leading American companies still rely more heavily on factory capacity, internal fleet output and expected future orders. This manufacturing advantage is not accidental; it reflects years of investment in robotics supply chains and production expertise.

Walden Robotics, despite being a Massachusetts-based Toyota spinout, illustrates how quickly deployment can happen when a company has access to Toyota's manufacturing infrastructure. The company says its general-purpose robots have been doing real production work, loading and unloading car parts, cleaning machinery and kitting for assembly, at a Toyota plant in North America since February 2026, moving from first pilot to production work in under two months. This speed of deployment suggests that manufacturing capability and customer access matter as much as raw technological innovation.

What Does Real-World Deployment Actually Look Like?

Beyond the funding announcements and production numbers, the sources reveal important details about what humanoid robots are actually doing in the field. Agility Robotics' Digit has accumulated its 65,000 operating hours across nine customer facilities, with several pilots progressing into continuing commercial agreements. This depth of deployment is more meaningful than raw shipment volume because it demonstrates that customers are willing to keep using the robots rather than returning them as expensive experiments.

Meanwhile, Alibaba's DAMO Academy has released RynnBrain 1.1, a 122-billion-parameter embodied foundation model that has already been deployed on three robot platforms, including Unitree G1, Astribot-S1 and Tianji Wuji. This model surpasses closed-source and open-source alternatives on spatial reasoning benchmarks, suggesting that software advances are enabling robots to perform more complex manipulation tasks. The convergence of better hardware, better software and actual customer deployment creates a virtuous cycle that could accelerate the transition from research to production.

What Happens Next for Unitree and Its Competitors?

Unitree's IPO proceeds will fund robot-model development, hardware research, new products and manufacturing capacity, according to standard IPO prospectus language. However, the company's stated intention to keep spending on research and development rather than treat the listing purely as an investor exit suggests management believes the market is still in an early growth phase. This contrasts with some mature tech IPOs where proceeds are used primarily for shareholder returns.

The broader competitive landscape remains fluid. Agility Robotics' narrow lead comes from the quality of its deployment hours rather than raw shipment volume. Unitree could move ahead by converting more of its volume into recurring industrial work. Figure AI could take the lead once its production ramp becomes a large paid fleet operating with minimal human intervention. None of these advantages is permanent, especially as hardware costs fall and autonomy models spread between platforms.

What makes the current leaders difficult to copy is that their advantages combine hardware design, software capabilities, manufacturing expertise, customer data and deployment experience. These elements reinforce each other in ways that are hard to replicate quickly. However, as the technology matures and becomes more commoditized, the competitive moats will likely narrow, creating opportunities for new entrants and shifting the balance of power in the market.