Unitree G1 Edu Ultimate E U7 Humanoid Robot with Revo 2 Basic Hands (G1EDU-U7)
In stock
- BRAND:
- UNITREE ROBOTICS
- MODEL:
- G1 EDU ULTIMATE E
- PART #:
- G1EDU-U7
- ORIGIN:
- China
- Warranty:
- 18 MONTHS
- AVAILABILITY:
- SUBJECT TO AVAILABILITY
- SKU:
- Unitree-G1-Edu-U7
Unitree G1 Edu Ultimate E U7 Humanoid Robot with Revo 2 Basic Hands (G1EDU-U7)
A humanoid robot that walks well but cannot pick anything up is a demonstration, not a research tool. The Unitree G1 Edu Ultimate E U7 Humanoid Robot with Revo 2 Basic Hands (G1EDU-U7) is built to close that gap: it pairs a compact bipedal body with five-finger hands, an onboard AI computer and a sensing package for mapping and depth, so a lab can start work on manipulation and perception without assembling the pieces itself.
The product is available through this page, by quotation.
Background
Unitree and the G1 family
Unitree Robotics is the manufacturer behind the G1, a compact humanoid platform that universities, research institutes and technology teams use to study walking, balance, manipulation and human-robot interaction. The G1 is not sold as a single machine. The versions named in the sources are the G1, the G1 EDU and six Ultimate configurations lettered A to F, with this product being the G1 EDU Ultimate E, also coded U7.
That lettering matters when you plan a purchase. Each Ultimate configuration is defined by what is fitted to the same basic body, most visibly the hands. The U7 carries two five-finger BrainCo Revo 2 Basic hands, so it sits among the dexterous-hand configurations rather than the entry-level ones. If you are still deciding how far up the range you need to go, the Unitree G1 category shows the whole family in one place.
Positioning and key technology
The U7 is a development platform. Its onboard computing is an NVIDIA Jetson Orin NX with 100 TOPS of AI performance, which means perception models and control code can run on the robot instead of on a tethered workstation. For sensing, it carries a Livox MID-360 3D LiDAR and an Intel RealSense D435i depth camera, a combination commonly used for mapping, localization and obstacle awareness.
The joints are driven by low inertia, high-speed internal rotor PMSM motors (permanent magnet synchronous motors), which give better response speed and heat dissipation. Connectivity is WiFi 6 and Bluetooth 5.2. Taken together, these choices describe a robot meant to be programmed, instrumented and modified, not simply switched on and watched.
Technical Specifications
| Specification | Value | Applies to |
|---|---|---|
| Hands | 2× five-finger BrainCo Revo 2 Basic hands, without tactile sensing | G1 EDU Ultimate E (U7) |
| Onboard computing | NVIDIA Jetson Orin NX, 100 TOPS AI performance | G1 EDU Ultimate E (U7) |
| Sensors | Livox MID-360 3D LiDAR and Intel RealSense D435i depth camera | G1 EDU Ultimate E (U7) |
| Maximum torque of knee joint | 120N.m | G1 EDU |
| Arm maximum load | About 3Kg | G1 EDU |
| Height, width and thickness (stand) | 1320x450x200mm | G1 EDU |
| Height, width and thickness (fold) | 690x450x300mm | G1 EDU |
| Weight (with battery) | About 35kg+ | G1 EDU |
| Single leg degrees of freedom | 6 | G1 EDU |
| Joint motor | Low inertia high-speed internal rotor PMSM (permanent magnet synchronous motor) | G1 EDU |
| Connectivity | WiFi 6, Bluetooth 5.2 | G1 EDU |
Specifications as published by the manufacturer; confirmed against the datasheet at quotation.
The manufacturer has not published the walking speed, the IP rating, the payload per hand or the operating temperature range; they are confirmed at quotation.
What do the Revo 2 Basic hands bring, and what do they leave out?
The hands are the most important part of this configuration. Two five-finger BrainCo Revo 2 Basic hands give the robot a human-shaped grasp, which is what you want when the research question is about picking up everyday objects, handing items over or learning manipulation from human demonstrations. A gripper with two jaws cannot answer those questions; a five-finger hand can at least ask them.
The word Basic carries a real limit. These hands are specified without tactile sensing, so the robot has no touch feedback from the fingertips in this configuration. Teams that depend on contact sensing for delicate grasping will want to plan around that, using vision and the arm and joint data instead, or by asking about other configurations. The arm load figure above describes the arm, not the hands.
If five fingers are not essential for your work, the Unitree G1 Edu Standard (G1EDU-U1) is listed on this site as a separate version, and the Unitree G1 Ultimate C (G1EDU-U5) is the other five-finger option named on the site. Ask for a side-by-side comparison with your request.
Why does the onboard computer matter?
Many humanoid projects stall at the point where perception, planning and control have to share one machine. The Jetson Orin NX with 100 TOPS of AI performance sits on the robot itself, so camera and LiDAR data can be processed where they are collected. That shortens the path from sensor to decision and removes a dependence on a cable or a heavy laptop following the robot around.
The sensing pair complements it. The Livox MID-360 is a 3D LiDAR for building a picture of the surroundings, while the Intel RealSense D435i is a depth camera for what is directly in front of the robot. With WiFi 6 and Bluetooth 5.2 available, a team can connect to the robot from a nearby workstation, push updates and collect logs. How much of that you use depends on your own software stack.
Who is the Unitree G1 Edu Ultimate E U7 for?
University and institute labs are the most natural buyers. A five-finger-handed humanoid with onboard AI computing and LiDAR lets a group study whole-body control, navigation and manipulation on one body, and then share that platform among students and projects.
Robotics companies and integration teams are a second group. They can use the U7 to prototype behaviors, test perception pipelines or build a demonstration for a customer, before deciding what a production design would need. Teaching programs in engineering and computer science also benefit, because the robot is a physical target for code that students would otherwise only see in simulation.
It is a less obvious choice if your work needs touch feedback at the fingertips, or if you only want a robot to walk and gesture. In the first case, ask us about other configurations; in the second, the Unitree G1 Edu Standard may fit better. Either way, tell us what the robot must do and we will help match the version to the job.
What is the Unitree G1 Edu Ultimate E U7 used for?
- Manipulation research. Study grasping, hand-over and object handling with a five-finger hand on a humanoid body.
- Navigation and mapping. Use the 3D LiDAR and depth camera to build maps and test localization methods.
- On-robot AI development. Run and refine perception and decision models on the Jetson Orin NX computer.
- Teaching and training. Give students a real humanoid on which to test control and programming exercises.
- Prototype demonstrations. Show customers or sponsors what a humanoid workflow could look like before committing to a larger program.
- Human-robot interaction studies. Explore how people respond to a robot of human proportions that can use its hands.
Pricing and Availability
We do not print prices on this page. Pricing is provided by quotation, and the quotation is where the exact configuration, accessories and your requirements are brought together. Delivery, customs and support are arranged with the quotation, so you know the full picture before you commit. Send us your request and the Unitree G1 Edu Ultimate E U7 Humanoid Robot with Revo 2 Basic Hands (G1EDU-U7) is available on this page.
Frequently Asked Questions
What hands does the U7 have?
It has two five-finger BrainCo Revo 2 Basic hands, without tactile sensing.
What computer is built into the robot?
An NVIDIA Jetson Orin NX with 100 TOPS of AI performance. It lets perception and control software run on the robot itself.
Which sensors does it carry?
A Livox MID-360 3D LiDAR and an Intel RealSense D435i depth camera, both for sensing the environment.
How large is the robot when standing and folded?
For the G1 EDU, the standing size is 1320x450x200mm and the folded size is 690x450x300mm. Its weight with battery is about 35kg+.
How does it connect?
The G1 EDU lists WiFi 6 and Bluetooth 5.2.
Do the hands have touch feedback?
No. The Revo 2 Basic hands in this configuration are specified without tactile sensing.
Where can I buy the Unitree G1 Edu Ultimate E U7 and what does it cost?
You can request it through this page. The cost is given by quotation, based on your configuration and delivery needs, and no price is shown here.
Summary
The idea behind the U7 is simple: a humanoid is only useful for research if it can reach, see and think on its own body. This configuration offers five-finger hands, a 100 TOPS onboard computer and LiDAR plus depth sensing on the compact G1 EDU platform. Its limit is just as clear: the hands have no tactile sensing. If that balance suits your project, this page is the starting point.
Specifications
General
Physical
Mobility
Performance
Computing
Sensors
Connectivity
Software
Battery + Power
Shipping
Feature
| Equipped with all the functions of G1-Edu standard version |
| Upgraded from 1 to 3 degrees of freedom in the waist. |
| Single arm freedom upgraded from 5 to 7, both arms upgraded. |
| Up to 29 DOF for whole robot |


