ShanghaiJune 24, 2026 /PRNewswire/ — From its founding to a valuation of $1 billion, Critical Point took just five months, becoming one of the fastest-growing unicorns in the embodied intelligence sector.
This funding round was co-led by strategic and industrial capital, including Mirae Asset, a major internet company, and Jingming Capital, with participation from industrial resource partners such as Dongfeng Asset, CITIC Digital, and Shanghai Electric Science and Technology Fund, as well as institutions including Fresh Capital, Kailian Capital, Puhua Capital, Tianji Capital, We Venture, Zhenyan Investment, and Mingjia Capital. Existing shareholders, including Hillhouse Ventures, BlueRun Ventures, and C Capital, continued to increase their investments.
The funds from this round will be concentrated in three areas: expanding mass production capabilities, building a dexterous manipulation data flywheel, and advancing the next phase of research and development for Contact Intelligence models.
The Other Side of Funding: A Company Profitable Since Inception
In the current embodied intelligence industry, the scale of financing is not the sole indicator of a company’s value. For investment institutions, the more critical question is whether a robotics company has already proven its product value and possesses sustained commercialization capabilities.
Critical Point’s profitability is built on actual product deliveries and market-driven customer demand.
The company’s business originates from the Zhiyuan ecosystem, with deep technical expertise, engineering, and manufacturing experience in dexterous hands. From its establishment in January 2026 to this funding round, Critical Point’s OmniHand series dexterous hands have cumulatively delivered over 8,000 units, and grippers have shipped over 10,000 units. This shipment volume ranks among the top in the domestic dexterous hand market in Q1 2026.
For dexterous hands, large-scale delivery means comprehensive validation across the supply chain, manufacturing processes, and quality control systems. From production line setup to mass production, Critical Point has accumulated systematic experience in process optimization and quality management during continuous capacity expansion. The automation level of production line testing is steadily improving, with full-process automated testing coverage expected by year-end. Product consistency and yield rates already support sustained large-scale deliveries. As production capacity ramps up and processes mature, overall manufacturing costs are entering a steady downward trajectory.
An even more noteworthy indicator than shipment volume is the customer structure. Critical Point’s current order sources cover multiple customer types, including humanoid robot manufacturers, industrial automation integrators, research institutions, and overseas developers. This means that Critical Point is actively chosen by customers from different industries and application scenarios based on their independent judgments in the open market. A diversified customer structure not only directly validates product competitiveness but also indicates higher stability and predictability for Critical Point’s future revenue growth.
As Robots Enter the Real World, the Competitive Focus is Shifting to “Manipulation Capability”
In recent years, breakthroughs in the embodied intelligence industry have primarily focused on the cognitive layer. Large models enable robots to understand environments, break down tasks, and generate action plans.
However, when robots truly enter industrial and real-world scenarios, simply “knowing what to do” is far from sufficient. Tasks such as grasping, assembly, and flexible manipulation require robots to continuously perceive changes and make precise adjustments during contact with the environment.
The core of the next phase of robotics competition is not just about enabling robots to “see the world,” but to understand the physical world through touch and force feedback. Critical Point defines this capability as “Contact Intelligence” and has engineered it through the DUET (Dual-layer Unified Embodied Tactile Intelligence) architecture.
The DUET architecture, through the synergy of the Maneuver Layer and the Contact Layer, allows robots to both plan actions based on task objectives and dynamically adjust manipulation strategies based on real-time tactile feedback, thereby improving operational stability in complex environments.
In June 2026, at ICRA in Vienna, Critical Point publicly validated the DUET dual-layer intelligence framework by completing a balloon dog folding task through the coordinated use of two dexterous hands. This demonstrated that when a robot possesses both long-sequence action capabilities and fine contact abilities, complex contact manipulation becomes possible.
From Dexterous Hands to the Gateway of Contact Intelligence: OmniHand 3 Ultra-M Explores Next-Generation Robot Manipulation Capabilities
The realization of Contact Intelligence relies not only on algorithmic models but also on the support of dexterous hand hardware for perception, control, and response capabilities. Focusing on this direction, Critical Point continues to push the performance boundaries of dexterous hands.
In June, the company released its latest generation fully direct-drive dexterous hand, the OmniHand 3 Ultra-M, further exploring the engineering balance between high degrees of freedom, miniaturization, and high-response control. Integrated within a structure close to the size of an adult palm, it features 20 active degrees of freedom, employs a fully direct-drive transmission architecture, and enhances tactile perception capabilities, enabling real-time sensing of contact state changes. This allows robots to move from merely “completing actions” to gradually understanding and adapting to the real physical environment.
In Critical Point’s technological roadmap, the dexterous hand is not just the robot’s end-effector but also a crucial gateway connecting the real world with intelligent models. In the future, the OmniHand 3 Ultra-M will also serve as an important platform for Contact Intelligence data collection and model training, continuously driving the iteration of robot manipulation capabilities.
Signal Significance: Embodied Intelligence is Ushering in a Commercialization Watershed
If Contact Intelligence explains why Critical Point has been able to stand out, then this funding round determines how fast the company can grow. Critical Point stated that the funds from this round will be primarily invested in three directions: stable mass production, the data flywheel, and Contact Intelligence model research.
- At the production level, building on the experience of delivering 8,000 units, Critical Point will continue to invest in improving yield rates, automated assembly, and engineering optimization for out-of-box consistency. For dexterous hands, transitioning from “being able to make them” to “being able to deliver them stably” is a crucial threshold for entering large-scale commercial deployment. The goal is to ensure that the 80,000th product maintains the same quality as the first, while continuously reducing manufacturing costs.
- At the data level, Critical Point is building its own data collection factory and will open-source parts of the dataset. This funding round will accelerate the factory’s construction pace and post-production expansion speed, allowing the data flywheel to enter full-speed operation as soon as possible.
- At the algorithm level, Critical Point will continue to invest in the research and development of Contact Intelligence models, further solidifying the underlying capabilities for dexterous manipulation.
These three investments are not independent of each other but together form a growth loop from product delivery and data accumulation to intelligent evolution.
The significance of Critical Point becoming a unicorn is not just another robotics company achieving a high valuation. More importantly, it provides a new observation sample: in the embodied intelligence industry, the capital market is beginning to focus not just on technology demonstrations, but on whether products can be delivered at scale, whether customers are willing to pay consistently, and whether the business model can form a positive cycle.
From this perspective, Critical Point’s 8,000-unit deliveries, first-quarter profitability, and continuously growing customer demand may be more noteworthy than the financing figures themselves.
The robotics industry is transitioning from “demonstrating feasibility” to “commercial viability.” And dexterous manipulation capabilities are becoming the key infrastructure in this process.