A1 is a budget-friendly bipedal robot optimized for reinforcement learning, capable of autonomously learning and refining its action policies in complex environments. Through generative motion, voice, emotion and content, it delivers multimodal human-robot interaction, and its capabilities keep growing with different Agents.
Automatic safety-stop mechanism and structural safety design, ideal for education and home use.
Learns autonomously and refines action policies in complex environments, supporting policy training and deployment.
Generative motion, voice, emotion and content make communication more natural and engaging.
Extends teaching, companionship, entertainment and more through different Agents, growing continuously.
| Height | 44.5 cm |
| Weight | 8 kg |
| Battery Life | 2 – 3 hours |
| Main Controllers | STM32, Odroid C4, Jetson Nano |
| RAM | 4 GB |
| Storage | 64 GB |
| ARM Network Inference Processor | Arm Cortex-A55 |
| Charging Port | DC |
| Ethernet Ports | 3 × RJ45 |
| Sensors | Customizable on request |
| Degrees of Freedom (DoF) | 10 DoF |
| Yaw & Ankle Torque | Rated 4N·m, peak 12N·m |
| Roll, Pitch & Knee Torque | Rated 8N·m, peak 20N·m |
| Max Speed | 0.5 m/s |
| Legged Walking | Supported |
| Automatic Safety Stop | Supported |
| Inertial Sensors | Gyroscope + Accelerometer |
| API / Research Programs | Secondary development SDK + RL training framework |
* Specifications are subject to change without prior notice due to product improvements or updates.
Contact us for education procurement plans, pricing, and curriculum partnership information.
