- New
Product description: Humanoid robot - Unitree G1 Edu U3
The Unitree G1 Edu U3 is a professional humanoid platform designed for research laboratories, technical universities, and teams developing modern AI and robotics systems. The robot combines advanced mechanics, advanced computational architecture, and precise manipulators enabling complex operations in real-world environments. The Edu U3 model is based on the expanded G1 Edu Advanced Edition platform and is equipped with two three-fingered Dex3-1 hands (without touch sensors) controlled by force . With a total of 43 degrees of freedom, the robot can perform very natural movements, maintain balance, and perform complex manipulation tasks requiring high precision. The robot utilizes a depth camera and a 3D LiDAR scanner for environmental perception, supports WiFi 6 and Bluetooth 5.2 , and a 9000 mAh intelligent battery provides approximately 2 hours of operation. The maximum knee joint torque is 120 Nm, while the arm can move objects weighing up to 3 kg . AI processing is handled by a powerful 8-core processor and the Nvidia Jetson Orin platform, offering up to 100 TOPS of computing power. The robot supports the development of proprietary algorithms, autonomy, and control systems and can be used for research on human-robot interaction, perception, motion planning, and object manipulation.

G1 Edu equipped with an Nvidia Jetson Orin compute unit, a Livox MID-360 3D LiDAR system and an Intel RealSense depth camera.
Main features of the Unitree G1 Edu U3 robot
- 43 degrees of freedom : very high flexibility of movement and natural humanoid kinematics
- Dex3-1 two hands : three-fingered force-controlled manipulators (no touch sensors)
- Advanced perception : depth camera and 3D LiDAR scanner
- AI computing power up to 100 TOPS : Nvidia Jetson Orin platform
- Wireless connectivity : WiFi 6 and Bluetooth 5.2
- Extensive programming capabilities : support for high-level and low-level development
- Precise PMSM drives : high dynamics and quick response of joints
- Advanced object manipulation : grabbing, moving, and manipulating elements
- OTA support : software and robot function updates
- Educational and research platform : ideal for R&D projects and AI development
Unitree G1 Edu Humanoid Robot Mechanics
The Unitree G1 Edu U3 utilizes a sophisticated humanoid design offering 43 degrees of freedom , enabling highly realistic reproduction of human movements and advanced manipulation. The robot can walk smoothly, maintain balance, rotate its torso, and perform complex movement sequences . Each leg has 6 degrees of freedom, while the arms and multi-axis torso enable natural movements of the entire body. Additionally, each Dex3-1 hand offers 7 degrees of freedom and force control, allowing for precise grasping and manipulation of objects. The design utilizes industrial-grade bearings and PMSM motors, ensuring high dynamics, fast response, and durability during intensive research and experimental work.

The U3 version uses the kinematics of the U2 version and additionally has two dexterous, three-fingered Dex3-1 hands.
Unitree G1 AI Robot Perception and Communication
The robot is equipped with an advanced perception system based on a depth camera and a 3D LiDAR scanner , enabling environmental analysis, spatial mapping, and real-time obstacle detection . This allows the platform to be used for research on autonomy and navigation. The audio system includes a four- microphone array and a 5W speaker, enabling voice communication and integration with speech recognition systems. The robot can participate in projects related to human-machine interaction and intelligent communication. The Unitree Explore APP allows you to monitor the robot's operating parameters, temperatures, alarms, and IMU data. You can also perform joint calibration and system diagnostics.
Movement dynamics and physical capabilities of the robot
The Unitree G1 Edu U3 has a compact design , measuring 1320 mm high and weighing approximately 35 kg . Thanks to powerful drives, the robot achieves a maximum knee torque of 120 Nm , which translates into stable and dynamic movement. The robot's arms can carry objects weighing up to approximately 3 kg , while the Dex3-1 hands enable precise manipulation. The robot can perform dynamic movements, maintain balance, and execute complex trajectories even while handling objects. The 9000 mAh battery provides approximately 2 hours of operation , and the quick-change system minimizes downtime during research and experiments.
Unitree G1 Edu U3 Robot Architecture, Programming, and Communication
The G1 Edu U3 platform was designed as an open research environment enabling the development of proprietary control, perception, and artificial intelligence systems . The robot utilizes an 8-core processor and the Nvidia Jetson Orin platform, providing computing power of up to 100 TOPS . The system supports the development of high- and low-level applications, enabling full control over the robot's behavior. The manufacturer provides technical documentation, APIs, training materials, and implementation examples. WiFi 6 and Bluetooth 5.2 connectivity allow for integration with computers, laboratory systems, and development environments. Over-the-air (OTA) updates enable convenient development of the platform's functionality without the need for physical intervention.
| UnitreeG1 Edu U3 Technical Specifications | |
|---|---|
| Total degrees of freedom | 43 |
| Degrees of freedom of the leg | 6 |
| Degrees of freedom of the waist | 3 |
| Degrees of freedom of the arm | 5 |
| Degrees of freedom of the hand | 7 (optional force control for three-fingered hand + 2 additional degrees of freedom for the wrist) |
| Dex3-1 Hand Parameters | thumb: 3 active degrees of freedom, index finger: 2 active degrees of freedom, middle finger: 2 active degrees of freedom |
| Optional installation of touch sensors | Yes |
| Joint bearings | industrial cross roller bearings with high precision and high load capacity |
| Joint motors | PMSM permanent magnet synchronous motors with a low-inertia internal rotor |
| Maximum knee joint torque | 120 N*m |
| Maximum load on the arm | about 3 kg |
| Calf and thigh length | 0.6 m |
| Arm span | about 0.45 m |
| Waist range of motion | Z ±155°, X ±45°, Y ±30° |
| Knee range of motion | from 0° to 165° |
| Hip range of motion | P ±154°, R from -30 to 170°, Y ±158° |
| Wrist range of motion | P ±92.5°, Y ±92.5° |
| Full internal cable management | Yes |
| Joint encoder | double encoder |
| Cooling system | local air cooling |
| Battery type | 13-cell lithium-ion battery |
| Basic computing power | High-performance 8-core processor |
| Sensors and communications | depth camera + 3D LiDAR |
| 4-microphone array | Yes |
| Loudspeaker | 5 in |
| Wireless connectivity | WiFi 6, Bluetooth 5.2 |
| High-performance computing module | Nvidia Jetson Orin |
| Quick-connect battery | Yes |
| Battery capacity | 9000 mAh |
| Charger | 54V / 5A |
| Manual controller | Yes |
| Working time | about 2 hours |
| OTA updates | Yes |
| Possibility of further development | Yes |
| Dimensions in standing position | 1320 x 450 x 200 mm |
| Dimensions when folded | 690 x 450 x 300 mm |
| Weight with battery | over 35 kg |
| G1 Edu Version Comparison | ||||
|---|---|---|---|---|
| Model | DOF | Hands | Touch sensors | |
| G1 Edu U1 | 23 | lack | NO | |
| G1 Edu U2 | 29 | lack | NO | |
| G1 Edu U3 | 43 | 2 x Dex3-1 | NO | |
| G1 Edu U4 | 43 | 2 x Dex3-1 | Yes | |
| G1 Edu U5 | 41 | 2 x RH56DFQ | NO | |
| G1 Edu U6 | 41 | 2 x RH56DFTP | yes - 17 sensors / hand | |
| G1 Edu U7 | 41 | 2 x Revo 2 Basic | NO | |
| G1 Edu U8 | 37 | 2 x Dex3-1 | NO | |
| G1 Edu U9 | 37 | 2 x Dex3-1 | yes - 33 sensors / hand | |
| G1 Edu U10 | 35 | 2 x Revo 2 Basic | NO | |
Useful links |



