Rapid growth
Reported revenue rose from roughly RMB159 million in 2023 to about RMB1.7 billion in 2025, while the business moved from losses to profitability.
Robotics / IPO Research
Unitree’s Public-Market Test
Profitable humanoid platform—or a high-margin hardware business whose industrial demand still needs proof?
Published August 13, 2026 · Evidence cut-off: August 13, 2026
Research posture
Watchlist initiation. Unitree has demonstrated meaningful revenue, profitability and hardware shipments, but this is not a completed valuation underwrite. At the evidence cut-off, primary sources confirmed CSRC registration approval and the launched STAR Market issuance process. This article does not treat a first trading date, ticker or final valuation as confirmed without an exchange listing announcement.
Core thesis
Unitree has proved that capable humanoid hardware can be manufactured, shipped and sold with attractive reported economics. It has not yet proved the more valuable proposition: that general-purpose humanoids can deliver reliable, profitable labor at industrial scale.
The valuation hinge is not whether the robots can perform impressive movements. It is whether research and early-adopter demand converts into repeatable factory, logistics and service deployments without compressing margins.
What the filings show
Unusual margins
Reported principal-business gross margin reached about 60% in 2025, supported by scale, procurement savings, product mix and component commonality.
Humanoid inflection
Humanoid robots rose to more than half of 2025 revenue, turning Unitree into a direct embodied-AI benchmark.
Mixed demand quality
Research and education represented most disclosed humanoid use, while industrial applications remained a small minority.
Hardware strength versus the “robot brain”
Unitree’s strongest evidence sits in actuators, mechanical design, motion control, cost engineering and production. But locomotion is not sufficient for economically useful general labor. The harder proof is autonomous perception, manipulation, long-horizon task completion, safety, uptime and workflow integration. The filings support a high-performance hardware thesis more strongly than software-like physical-AI economics.
Public-market read-throughs
| Name | Comparison | What matters |
|---|---|---|
| TSLA | Unitree has stronger external hardware-sales evidence; Optimus targets a broader AI, data and internal-deployment flywheel. | Autonomous deployments and robot economics. |
| VPG | Higher volumes can expand precision-sensing demand, but supplier content is unproven. | Design wins and content per robot. |
| TER | Unitree validates robotics manufacturing activity; TER remains diversified through automation and semiconductor test. | Robot-related orders. |
| SYM | General robots may challenge specialized automation, but warehouse uptime currently favors purpose-built systems. | Customer ROI and reliability. |
Unitree is not in the AleaBito Robotics USD Index. Inclusion would require a verified listing, accessible historical closes and an explicit rebalance decision.
Valuation and falsifiers
Unitree can be framed as a premium hardware maker, developer platform, automation company or physical-AI platform. A defensible valuation must separate current hardware earnings from future platform optionality and wait for verified offer terms, diluted shares, float and trading price.
- Proof: industrial-use revenue, repeat orders, uptime and labor ROI, sustained margins after service costs.
- Warning: shipments remain concentrated in research, education or demonstrations while costs rise.
- Break: industrial deployments stall, pricing falls faster than costs, or autonomy requires persistent teleoperation.
Sources
Related research
Robotics USD Index · TSLA Optimus analysis · Robotics stock-picking methodology