Summary of Key Points
The “dexterous hand” of humanoid robots represents the most critical and challenging aspect in the industry to overcome—its cost accounts for 10%-20% of the robot’s overall cost (Tesla’s Optimus model has a cost of 17.3%), which determines whether the robot can perform delicate tasks such as peeling oranges or holding smartphones. Capital is pouring into this field: domestic financing reached 16.877 billion yuan in 2025, and it exceeded 25 billion yuan in the first half of 2026, with leading companies seeing their valuations soar (Lingxin Qiaoshou’s target valuation is 6 billion US dollars). However, the industry is currently in a state of “loss of focus” due to various issues: chaotic technical approaches (full direct drive/rope drive/linkage systems), inconsistent metrics for shipment volume, degrees of freedom, and a lack of standards. In the future, leading robot manufacturers developing their own components combined with the breakthrough of tactile technologies will trigger a reshuffle in the market, shifting the competition from focusing on raw parameters to implementing solutions in real-world scenarios.
1. The Dexterous Hand: The Robot’s Lifeline—High Cost and Complexity
Why is the dexterous hand so important? Consider these two facts:
- High cost percentage: The material cost of Tesla’s Optimus’ dexterous hand accounts for 17.3%, and in higher-end models, it can exceed 20%, making it more expensive than the robot’s “legs” or “torso”.
- Function determines value: Ordinary grippers can only move boxes, but a dexterous hand can perform precise tasks like screwing screws or peeling oranges; without it, the robot is essentially useless and unable to perform meaningful work.
The main challenges lie in motors and fine control: The full direct drive approach requires an independent motor for each joint, which not only results in high costs (each hand motor costing between 26,000 and 65,000 yuan) but also relies on imports. Rope drive and linkage systems are cheaper but lack flexibility and are unable to perform complex movements.
2. Three Technological Approaches Favored by Capital—Each with Its Own Niche
The dexterous hands available on the market vary significantly, and the chosen technology determines their target audience and potential profit:
- Full direct drive: Each joint is driven by an independent motor, offering the highest level of flexibility (imitating human hand movements), but the cost starts from 100,000 yuan. The main customers are research institutions; these robots may eventually enter households, with the highest valuations (Lingxin Qiaoshou’s valuation at 6 billion US dollars). However, they have limited load-bearing capacity due to the low torque of small motors and rely on imports.
- Rope drive: Motors are placed at the wrist, and ropes are used to move the fingers; these systems are very inexpensive (some prototypes cost less than 1,000 yuan), but they offer limited flexibility and are suitable for simple grasping tasks (e.g., moving boxes), targeting university laboratories for teaching demonstrations.
- Linkage system: Motors are located near the fingers, and linkage mechanisms are used to control the joints; these systems cost between 10,000 and 50,000 yuan. They have stronger load-bearing capacity but lower precision, making them suitable for factories that need to replace manual labor with repetitive tasks.
A hybrid drive (combining rope and direct drive) offers a balance of cost and flexibility. However, capital prefers the full direct drive approach due to its potential for widespread adoption in households.
3. The Industry’s Lack of Focus: Confusion and Standards
The industry is currently in a state of disarray without unified standards:
- Inconsistent shipment figures: Different companies report different numbers (e.g., one company claims 10,000 units shipped; another says it occupies 80% of the high-degree-of-freedom market); there are also discrepancies in how degrees of freedom are calculated.
- Varying cost ranges: Prices range from 3,999 yuan (entry-level rope drive models) to several hundred thousand yuan (high-end direct drive models), with a 25-fold difference.
As a result, the company with the highest shipment volume may not necessarily generate the most revenue (selling 10,000 units at 3,999 yuan is less profitable than selling 2,000 units at 100,000 yuan each). Investors are more interested in the “future potential” of these technologies rather than immediate sales figures.
4. The Future of the Industry: Shifting from Competition Based on Parameters to Implementation in Real Scenarios
Two factors will transform the industry:
- Leading robot manufacturers developing their own components: Companies like Tesla and Yushu have already started developing dexterous hands independently. Zhiyuan, a recently split company, raised four rounds of funding within five months (with a valuation of 1 billion US dollars). While developing their own components helps improve compatibility with various applications, large-scale production still requires cooperation with third parties.
- Breakthrough in tactile technologies: Tactile sensors are becoming standard rather than optional, with installation rates increasing from 10% in 2024 to over 50% in 2025. Robots without tactile sensors cannot perform delicate tasks accurately; those with them can handle complex operations like peeling oranges.
Future trends:
- Short term (1-2 years): Multiple technical approaches will coexist, and purchasing pre-made components will remain the norm for smaller manufacturers due to high costs.
- Medium term (3-5 years): Smaller manufacturers will be phased out, and leading third-party suppliers will collaborate with robot manufacturers.
- Long term: The competition will focus on implementing solutions in real-world scenarios—not just on the number of degrees of freedom (which already exceed human hands’ 27), but on which technology can generate actual profits in practical applications such as surgery or agriculture.
The industry does not lack dexterous hands; what’s needed are robots that can effectively perform tasks. The company that first completes the cycle of “technology → application → data” will emerge as the winner.
(The translation maintains the original Markdown structure, using clear and natural language suitable for financial journalism, adapting expressions to fit the target audience.)