EngineAI T800: Confirmed Specs, Demos and Limits
The T800’s athletic demonstrations show powerful motion, while autonomy, safety and commercial specifications remain less transparent.
By TechniaHQRobot
A measured technical profile of EngineAI’s T800 humanoid, separating reported height, mass, joints, battery and demonstrations from unverified claims.
EngineAI is a Shenzhen humanoid robotics company founded in 2023.
Public reports describe T800 as approximately 1.73 meters tall, 75 kilograms and 29 degrees of freedom.
Demonstrations include running, kicks and acrobatic motion.
The company has not published enough English technical material to verify every sensor, compute and battery claim independently.
Original X post
Open on XT800 is positioned as a full-size general-purpose humanoid
EngineAI introduced T800 after earlier biped and humanoid platforms. Public reports describe a full-size machine around 1.73 meters tall and 75 kilograms with 29 degrees of freedom. The frame is reported to use magnesium-aluminum alloy.
These figures are useful as manufacturer-reported specifications. A complete official English datasheet was not located, so details should not be compared with competitors as independently certified measurements.
Athletic demos emphasize actuator power and balance
T800 videos show running, roundhouse kicks, capoeira-style motion and impact demonstrations. Such actions demand fast joint control, high leg torque and recovery after momentum shifts.
A viral video of the robot kicking EngineAI founder Zhao Tongyang was released after viewers questioned whether earlier footage was computer generated. The demonstration shows a physical robot producing significant force. It does not establish safe operation near unprotected people.
Technical details
- Robot
- EngineAI T800
- Developer
- EngineAI Robotics, Shenzhen
- Reported height
- Approximately 1.73 m
- Reported mass
- Approximately 75 kg
- Reported joints
- 29 degrees of freedom
- Reported battery
- Modular solid-state lithium system; up to four hours claimed
- Control status
- Not disclosed clearly for every demonstration
- Availability
- Public reservation and deployment claims require official confirmation
The control method remains unclear
Public coverage does not consistently state whether each motion was autonomous, scripted, motion-captured or teleoperated. A preprogrammed athletic sequence is still a difficult control achievement, but it is different from choosing and executing an action in an unfamiliar environment.
EngineAI should publish the controller mode, sensor inputs and number of attempts for each demo. That information would let viewers separate hardware capability from task intelligence.
Battery and sensing claims need primary documentation
Reports cite a modular solid-state lithium battery with up to four hours of operation, along with cameras and environmental sensing. Runtime depends heavily on standing, walking, manipulation and high-power athletic motion.
Compute hardware, sensor models and sustained thermal limits are not clearly documented in an accessible official source. The article therefore keeps them as reported claims rather than verified specifications.
Commercial usefulness will depend on safer tasks
EngineAI has described logistics, hospitality and industrial work as possible applications. Those tasks require reliable grasping, navigation, fleet management and safety around workers. Kicking strength is not a direct measure of usefulness in any of them.
The next evidence should include a named task, autonomous or supervised control label, success rate, runtime and maintenance data. T800 is a real and capable motion platform, but broad commercial readiness remains unproven.
Verification notes
- Height, mass, degrees of freedom and battery figures are identified as reported specifications because an authoritative full English datasheet was not available.
Explore the technical topic
Share this article
Share the current TechniaHQRobot article page.
Sources
Editor : @techniahqrobot
TechniaHQRobot editorial coverage on AI, robotics, automation and Physical AI.