Xynova's Prima 1 Dexterous Hand Takes a CES 2027 Innovation Award

The 22-degree-of-freedom direct-drive hand won in the Artificial Intelligence category and will demo live at booth 12918.
Xynova's Prima 1, a fully direct-drive dexterous hand with 22 degrees of freedom, has been named a CES 2027 Innovation Awards Honoree in the Artificial Intelligence category. The Hangzhou company announced the honor on November 10, two months before CES 2027 opens in Las Vegas on January 6.
The hand is built as an open research platform for embodied AI. Its direct-drive architecture answers in milliseconds, with precise force control, low backdrive force and high force transparency. Five-finger visuo-tactile sensing captures contact feedback for data collection and manipulation research, and the hand is rated for continuous, long-duration operation. On Xynova's own numbers, a fingertip delivers 20 newtons of force, the hand lifts up to 20 kilograms, and a grip closes in 0.1 seconds.
Prima 1 supports imitation learning, teleoperation and Sim-to-Real policy transfer, with SDK access to control functions and operating data so researchers can build and test their own manipulation models. “Dexterous manipulation is what will let humanoid robots move beyond basic tasks and interact meaningfully with the physical world,” said Alex Xia, founder and CEO of Xynova. “Being named a CES 2027 Innovation Awards Honoree marks an important milestone for Xynova and reinforces our commitment to advancing physical AI.”
At CES 2027, Xynova will show Prima 1 alongside Flex 2, a hybrid-drive dexterous hand built for precision manipulation in real-world applications. Both will run live demonstrations at booth 12918 in the North Hall of the Las Vegas Convention Center. The award puts a second dexterous hand on the CES 2027 honors list: AGIBOT's OmniHand 3 Ultra was honoured in Robotics and Artificial Intelligence when the awards were revealed on November 9. Xynova, a full-stack dexterous manipulation provider headquartered in Hangzhou with a 5,400-square-meter manufacturing facility, is also exhibiting at the Conference on Robot Learning this week and at Web Summit in Lisbon.