Key Products:

• Booster T1: A versatile humanoid robot featuring 23 degrees of freedom, equipped with a 14-core processor and NVIDIA AGX Orin GPU delivering 200 TOPS of AI performance. It supports ROS2, offers open-source SDKs, and is ideal for research, development, and RoboCup competitions.

• Booster K1: A compact humanoid robot designed for education and interactive demonstrations. It combines mobility, computing power, and ergonomics in a ready-to-use format, making it suitable for STEM education and research laboratories.

Platform lineup

• Booster T1 – a mid-size humanoid aimed at labs and competitions, with approximately 1.18–1.2 m height, ~30 kg mass, and 23 degrees of freedom (DoF) in the standard configuration. The design balances mobility, robustness, and extensibility (e.g., grippers or dexterous hands), making it suitable for tasks like sports demonstrations and manipulation research.

• Booster K1 – a compact “introductory embodied development platform” standing ~95 cm and ~19.5 kg, designed to be portable for classrooms, labs, and exhibitions. K1 emphasizes developer ergonomics (carryability, quick setup) and ships with multiple onboard compute options. 

Developer-first design

Both platforms are marketed as “lightweight, flexible, and tough,” with fall-tolerant structures and plug-and-play modules. The company publishes open documentation and an SDK to encourage secondary development, simulation, and integration with common AI pipelines. 

Competition pedigree and partnerships

Booster frequently showcases its robots in RoboCup-related scenarios and other public events (including an NVIDIA event where the T1 was featured). These activities position the systems as proven in fast-paced, real-world trials such as robot football and agility challenges. 

Technology and Specifications

Booster T1 (representative specs)

• Dimensions & mass: ~1.18–1.2 m tall; ~30 kg.

• Degrees of freedom: 23 DoF baseline (legs 6×2, arms 4×2, waist 1, head 2). Optional packages add grippers or dexterous hands for higher total DoF.

• Compute: NVIDIA Jetson AGX Orin configurations up to ~200 TOPS for on-robot perception and control. 

• Battery & runtime: vendor literature cites ~2 h walking / ~4 h standing in RoboCup scenarios. 

• Sensing & I/O: depth vision, mic array/speaker options, Wi-Fi 6/Bluetooth 5.2/optional 5G, and standard ports for lab integration.

Booster K1 (representative specs)

• Form factor: ~95 cm height; ~19.5 kg weight. Degrees of freedom: 22 DoF, aimed at agility and “sports-like” demos via Booster GYM 2.0.

• Compute tiers: published options at 48 / 117 / 200 TOPS, allowing developers to match budget to workload (vision, RL inference, whole-body control). 

Software, SDK, and open materials

Booster provides manuals (T1/K1), SDKs, and open-source resources for motion, perception, and integration. This documentation supports quick starts, secondary development, and competition prep, reflecting the firm’s “built for developers” ethos. 

Applications and Use Cases

Education and research

Universities and research labs use Booster platforms for legged locomotion, whole-body control, embodied AI, and robot manipulation studies. The smaller K1 lowers the barrier to entry for classroom demonstrations and student projects, while T1 provides headroom for higher-torque, longer-duration experiments. 

Competitions and benchmarking

Booster systems appear in RoboCup-style football scenarios and public trials that stress autonomy, perception, and agility. Such events supply benchmark tasks (walking, dribbling, falls and recoveries), helping teams measure progress across software stacks and mechatronic configurations. 

Embodied AI prototyping

With onboard Orin compute and documented interfaces, developers prototype policy inference, vision-based navigation, and learning from demonstration on the robot—then iterate quickly using SDK tools and simulation content.

Public engagement and demos

K1’s portability and T1’s higher payload/DoF options make them suitable for trade shows, outreach, and interactive exhibitions, where safe, repeatable motions and resilience to bumps/falls matter. 

Advantages / Benefits

• Developer accessibility. Clear manuals, SDK, and open materials reduce setup friction and time to first experiment—valuable for semester timelines and hackathon-style research sprints. 

• Configurability. From base locomotion to manipulation packages (grippers or higher-DoF hands), teams can scale capability as needs grow. 

• Performance-per-kilogram. T1’s ~30 kg and K1’s ~19.5 kg profiles strike a balance between safety, transportability, and useful torque/compute for embodied tasks. 

• Competition-tested. Appearances in RoboCup contexts and public events validate robustness and help mature the software ecosystem through community feedback. 

FAQ Section

What is Booster Robotics?
Booster Robotics is a Beijing-based company (founded in 2023) that develops humanoid robots and publishes SDKs and manuals aimed at researchers and educators. 

How does Booster’s humanoid platform work?
The robots combine legged locomotion, torso/arm/head joints (e.g., T1’s 23-DoF baseline) and onboard NVIDIA Orin compute (up to ~200 TOPS) for perception and control, with APIs for secondary development. 

Why is Booster important to embodied AI?
By offering accessible, robust hardware plus an SDK and documentation, Booster lowers the barrier for labs and classrooms to test learned policies, teleop pipelines, and perception stacks on real hardware.

What are the benefits of T1 vs. K1?
T1 offers more headroom (size, DoF options, runtime tiers) for ambitious locomotion/manipulation research; K1 prioritizes portability (≈95 cm/19.5 kg) and lower entry cost for education and demos.

Summary

Booster Robotics targets the developer and research segment of humanoid robotics with two accessible, open platforms. T1 and K1 emphasize durability, portability, and documented interfaces for rapid experimentation in locomotion, manipulation, and embodied AI. Public competition pedigree and an expanding documentation stack make Booster a practical choice for labs, classrooms, and teams that value fast iteration, reproducibility, and community-visible benchmarks.