AgiBot X2-N vs Green by Sberbank

Green is a full-size humanoid at 180 cm and 75 kg, with 41 degrees of freedom, including 12 finger joints and 4 wrist joints. AgiBot X2-N is a smaller platform built around mobility, with foot wheels added to a bipedal body. Green's design centres on its hands; the X2-N's on its legs.

The X2-N is the more versatile mover: it walks, rolls on its foot wheels, climbs stairs and handles slopes, and in AgiBot's demo it carried about 6 kg while climbing stairs. Green walks at up to 0.9 m/s and navigates unfamiliar indoor spaces on its own, steering around obstacles.

This is the sharpest contrast. The X2-N has no cameras at all and feels its way using joint feedback, foot pressure and an IMU. Green sees with eight RGB cameras and two LiDARs and senses force in its fingertips, which it needs for recognising and handling objects.

The X2-N's intelligence is focused on locomotion: it chooses its gait and switches between walking and rolling from what its body feels. Green's AI is focused on tasks and people, combining Sber's GigaChat for conversation with the Green-VLA vision-language-action model for manipulation, and it can correct its own mistakes.

Neither maker has published full battery figures. Sber says Green uses a 48 V battery, without runtime or charging time. AgiBot has not published the X2-N's own runtime, but it shares the AgiBot X2's battery platform, which runs for about 2 hours.

AgiBot positions the X2-N for logistics, healthcare and public services, where getting around mixed terrain matters. Green targets hands-on work: collecting items and stocking shelves, assembling and packing products, handling test tubes and bio samples, and room service and cleaning.

The X2-N runs AgiBot's Linux and ROS-based software. Green runs on Sber's GigaChat and the open-weight Green-VLA model, and Sber's Green Challenge lets outside teams train controllers in a simulator of the robot. Sber has not announced a public SDK.

The X2-N adjusts its gait in real time and has an emergency stop. Green relies on inertial and force sensors for balance, avoids obstacles and can detect its own errors; Sber has not published safety certifications. Ask both makers for certification details before deploying near people.