Entry-level prices are now below 4,000 yuan, with mainstream models under 10,000 yuan as self-developed actuators and volume production reshape the market.
Five years ago, a high-degree-of-freedom robotic hand could cost more than 1 million yuan ($149,700), making it an expensive laboratory instrument. Today, entry-level Chinese models are available for as little as 3,900 yuan, while mainstream products have fallen below 10,000 yuan.
The dramatic price cuts reflect more than aggressive discounting. They stem from domestic manufacturers developing their own core components, expanding production and offering products tailored to different applications. The result is a rapidly changing market for one of humanoid robots’ most complex components.
Replicating the human hand’s ability to grasp, pinch and manipulate objects requires much more than a collection of motors. A high-degree-of-freedom robotic hand can contain more than a dozen miniature servo actuators — devices that convert electrical energy into precise mechanical movement — alongside multidimensional tactile sensors, precision transmission mechanisms and metal structural components. Assembly and calibration are labor-intensive, limiting production capacity and yields.
The costs once put these devices beyond the reach of most buyers. The Shadow Hand, made by British company Shadow Robot, has historically been priced at around 1.5 million yuan, with some high-end models costing 1.6 million to 2.6 million yuan. The research-oriented Allegro Hand has cost about $16,000, equivalent to roughly 110,000 yuan.
Beijing Inspire-Robots Technology, one of China’s earliest commercial producers, also initially charged more than one million yuan per hand, selling only one or two units a year.
Over the past five years, however, manufacturers have pursued a two-track strategy: affordable models for volume sales and premium products for demanding applications.
Price wars
Inspire-Robots illustrates the scale of the shift. Its prices fell from more than 1 million yuan initially to 100,000 yuan when it launched China’s first commercially mass-produced dexterous hand in 2020, then to 50,000 yuan in 2023 and to around 20,000 yuan in 2024-25. The company has even cited a target price of 10,000 yuan.
The reductions have been supported by in-house development of miniature servo electric cylinders and integrated joints, two of the hand’s most expensive components. Its Suzhou manufacturing base has annual capacity of 500,000 electric cylinders and about 50,000 dexterous hands.
LinkerBot has taken product segmentation even further. Its L20 and L30 models, aimed at research and high-end industrial applications, cost 50,000 yuan to 99,900 yuan — roughly one-twentieth the price of comparable Shadow Hand products. At the lower end, its O7 costs 8,800 yuan, the O6 6,600 yuan and the O6 Lite as little as 3,900 yuan after subsidies. The 370-gram O6 can handle a load of 50 kilograms.
Humanoid robot makers and other industrial players are also entering the market. AgiBot launched its OmniHand entry-level model in August 2025 with a recommended price of 14,800 yuan. Orbbec Robotics’ ROH-LiteS001 targets industrial and consumer applications at 8,888 yuan, with annual production capacity approaching 10,000 units at its Zhuhai factory.
BrainCo’s Revo1 basic model costs around 35,000 yuan and targets rehabilitation and industrial applications. Meanwhile, PaXini’s high-end DexH13, equipped with advanced tactile sensing, remains priced at 98,000 yuan. Its fourth-generation PX6AX GEN4 multidimensional tactile sensor, however, has fallen to just 199 yuan, highlighting the falling cost of underlying components.
Five forces behind the price cuts
Several factors are driving the decline in prices.
First, domestic manufacturers are developing their own miniature actuators, precision transmission systems, and tactile sensors, reducing dependence on expensive imported components.
Second, product segmentation allows companies to simplify lower-priced models, removing unnecessary complexity while retaining more advanced sensing and motion capabilities in premium versions.
Third, humanoid robot makers are pressuring suppliers to cut costs. If a robot’s hand costs more than the rest of its body can economically justify, mass commercialization becomes difficult.
Fourth, automated production lines are improving manufacturing efficiency and yields, replacing labor-intensive small-batch assembly.
Finally, intensifying competition and investment are encouraging aggressive pricing as companies seek market share and relationships with major robot manufacturers.
Price is no longer the only measure
Further reductions are likely for lightweight, lower-degree-of-freedom hands as domestic actuator production expands. Models costing a few thousand yuan could become the main volume segment, although their capabilities remain limited, particularly for complex manipulation and fine-motor tasks.
High-degree-of-freedom hands with comprehensive tactile sensing are unlikely to see similarly dramatic price falls in the near term. Precision sensors, intricate transmission mechanisms and demanding assembly and calibration requirements keep costs high. Research institutions and advanced manufacturers are likely to remain their main customers.
Price is also becoming less important than reliability. Operating life, environmental adaptability and long-term stability will increasingly shape purchasing decisions. Some inexpensive models still have shortcomings in sustained operation, and industry participants have questioned whether prices of just a few thousand yuan are economically sustainable without sacrificing profitability.
Falling component costs could nevertheless create further room for reductions. The decline in tactile sensor prices from tens of thousands of yuan to 199 yuan illustrates how cheaper underlying technology can lower the cost of complete robotic hands.
The transition from million-yuan laboratory equipment to mass-produced components costing a few thousand yuan marks an important step towards commercializing humanoid robots. But lower prices do not eliminate the engineering challenges.
The next phase of competition will depend less on who can make the cheapest hands and more on who can deliver reliable, durable hands capable of handling the complexity of real-world tasks.
