Unitree Robotics

With a height of 64.5cm (standing), 33cm (prone), the B2 is designed to operate at a mid-range level — suitable for navigating under tables, around furniture, and through standard doorways without issue. This compact-but-capable size balances visibility with maneuverability.

Weighing 60kg (battery included), the B2 is a substantial machine. This weight provides stability during physical tasks and manipulation but means it requires careful consideration for floor loading and may need dedicated charging infrastructure. Industrial-weight robots typically offer higher payload capacity and more robust construction.

The B2 offers 4–6 hours (unloaded walking >5h / 20km; 20kg load >4h / 15km) of battery life per charge. Battery life is one of the most critical real-world performance metrics for any mobile robot. It determines how much work the robot can accomplish in a single session before needing to recharge. For quadruped robots, this runtime should be evaluated against the size of the area you need covered and the intensity of the tasks involved. Robots with self-charging capability can partially compensate for shorter battery life by autonomously returning to their dock.

The B2 can move at up to 6+ m/s (21.6 km/h). Maximum speed affects how quickly the robot can traverse its operating area, respond to commands, and complete tasks. For quadruped robots, speed must be balanced against safety — faster robots need better obstacle detection and stopping capabilities to prevent collisions and ensure safe operation around people and pets.

The B2 runs on Intel Core i5/i7 + optional Jetson Orin NX (up to 3 compute units) for its artificial intelligence capabilities. The AI platform determines how intelligently the robot behaves — from basic reactive responses to sophisticated scene understanding, natural language processing, and adaptive learning. A more advanced AI platform generally means better obstacle avoidance, more natural interaction, and the ability to improve performance over time through software updates.

With a height of 63.6cm (standing), 29.7cm (folded), the B1 is designed to operate at a mid-range level — suitable for navigating under tables, around furniture, and through standard doorways without issue. This compact-but-capable size balances visibility with maneuverability.

The B1 weighs About 50kg (including battery); battery about 5kg. Weight affects stability, portability, floor compatibility, and how the robot interacts with its environment.

The B1 offers 2h continuous walking / 5h standing (unloaded) of battery life per charge. Battery life is one of the most critical real-world performance metrics for any mobile robot. It determines how much work the robot can accomplish in a single session before needing to recharge. For quadruped robots, this runtime should be evaluated against the size of the area you need covered and the intensity of the tasks involved. Robots with self-charging capability can partially compensate for shorter battery life by autonomously returning to their dock.

The B1 requires 1–2 hours to reach a full charge. Charging time directly impacts the robot's daily operating capacity — faster charging means less downtime and more productive hours. Combined with its battery life, the charge-to-runtime ratio reveals how much of each day the robot can actually spend working versus sitting on its dock.

The B1 runs on Intel Core i5-1135G7 + Jetson Xavier NX ×3 for its artificial intelligence capabilities. The AI platform determines how intelligently the robot behaves — from basic reactive responses to sophisticated scene understanding, natural language processing, and adaptive learning. A more advanced AI platform generally means better obstacle avoidance, more natural interaction, and the ability to improve performance over time through software updates.

At just 40cm (standing) tall, the Go2 has a compact form factor that allows it to navigate under furniture, access tight spaces, and maintain a low profile during operation. Compact robots are particularly effective for cleaning, surveillance, and utility tasks.

The Go2 weighs ~15kg. Weight affects stability, portability, floor compatibility, and how the robot interacts with its environment.

The Go2 offers 1–2h (standard) / 2–4h (EDU long endurance) of battery life per charge. Battery life is one of the most critical real-world performance metrics for any mobile robot. It determines how much work the robot can accomplish in a single session before needing to recharge. For quadruped robots, this runtime should be evaluated against the size of the area you need covered and the intensity of the tasks involved. Robots with self-charging capability can partially compensate for shorter battery life by autonomously returning to their dock.

The Go2 can move at up to Air: 0-2.5 m/s; Pro: 0-3.5 m/s; X/EDU: 0-3.7 m/s (max ~5 m/s in lab). Maximum speed affects how quickly the robot can traverse its operating area, respond to commands, and complete tasks. For quadruped robots, speed must be balanced against safety — faster robots need better obstacle detection and stopping capabilities to prevent collisions and ensure safe operation around people and pets.

The Go2 runs on 8-core CPU, AI simulation-trained gaits, optional NVIDIA Jetson Orin (40–100 TOPS) for its artificial intelligence capabilities. The AI platform determines how intelligently the robot behaves — from basic reactive responses to sophisticated scene understanding, natural language processing, and adaptive learning. A more advanced AI platform generally means better obstacle avoidance, more natural interaction, and the ability to improve performance over time through software updates.

At 182cm, the Unitree H2 is roughly the height of an average adult human, which allows it to interact naturally with human-designed environments including countertops, doorways, and shelving at standard heights. This size is important for robots that need to work alongside people in factories, warehouses, or homes.

The Unitree H2 weighs ~70kg. Weight affects stability, portability, floor compatibility, and how the robot interacts with its environment.

The Unitree H2 offers About 3 hours of battery life per charge. Battery life is one of the most critical real-world performance metrics for any mobile robot. It determines how much work the robot can accomplish in a single session before needing to recharge. For humanoid robots, this runtime should be evaluated against the size of the area you need covered and the intensity of the tasks involved. Robots with self-charging capability can partially compensate for shorter battery life by autonomously returning to their dock.

The Unitree H2 runs on Up to 2070 TOPS (Jetson AGX Thor optional); Intel Core i5/i7 onboard for its artificial intelligence capabilities. The AI platform determines how intelligently the robot behaves — from basic reactive responses to sophisticated scene understanding, natural language processing, and adaptive learning. A more advanced AI platform generally means better obstacle avoidance, more natural interaction, and the ability to improve performance over time through software updates.

At just 45.7cm (standing) tall, the As2 has a compact form factor that allows it to navigate under furniture, access tight spaces, and maintain a low profile during operation. Compact robots are particularly effective for cleaning, surveillance, and utility tasks.

The As2 weighs ~18kg (with battery). Weight affects stability, portability, floor compatibility, and how the robot interacts with its environment.

The As2 offers AIR: ~2 hours unloaded (~10km), >1.5 hours with 10kg load (~7km); PRO: ~4 hours unloaded (~20km), >2.5 hours with 13kg load (~13km); EDU: ~4 hours unloaded (~20km), >2.5 hours with 15kg load (~13km) of battery life per charge. Battery life is one of the most critical real-world performance metrics for any mobile robot. It determines how much work the robot can accomplish in a single session before needing to recharge. For quadruped robots, this runtime should be evaluated against the size of the area you need covered and the intensity of the tasks involved. Robots with self-charging capability can partially compensate for shorter battery life by autonomously returning to their dock.

The As2 can move at up to 5+ m/s (EDU); 3.7 m/s (PRO); 3.0 m/s (AIR). Maximum speed affects how quickly the robot can traverse its operating area, respond to commands, and complete tasks. For quadruped robots, speed must be balanced against safety — faster robots need better obstacle detection and stopping capabilities to prevent collisions and ensure safe operation around people and pets.

The As2 runs on 8-core high-performance CPU + optional NVIDIA Jetson Orin NX (EDU) for its artificial intelligence capabilities. The AI platform determines how intelligently the robot behaves — from basic reactive responses to sophisticated scene understanding, natural language processing, and adaptive learning. A more advanced AI platform generally means better obstacle avoidance, more natural interaction, and the ability to improve performance over time through software updates.

With a height of 123cm (standing), the R1 is designed to operate at a mid-range level — suitable for navigating under tables, around furniture, and through standard doorways without issue. This compact-but-capable size balances visibility with maneuverability.

The R1 weighs ~29kg (with battery; ~27kg for Air model). Weight affects stability, portability, floor compatibility, and how the robot interacts with its environment.

The R1 offers ~1 hour (mixed activity) of battery life per charge. Battery life is one of the most critical real-world performance metrics for any mobile robot. It determines how much work the robot can accomplish in a single session before needing to recharge. For humanoid robots, this runtime should be evaluated against the size of the area you need covered and the intensity of the tasks involved. Robots with self-charging capability can partially compensate for shorter battery life by autonomously returning to their dock.

The R1 runs on 8-core CPU + GPU; optional NVIDIA Jetson Orin (40–100 TOPS, EDU only); UnifoLM multimodal LLM for its artificial intelligence capabilities. The AI platform determines how intelligently the robot behaves — from basic reactive responses to sophisticated scene understanding, natural language processing, and adaptive learning. A more advanced AI platform generally means better obstacle avoidance, more natural interaction, and the ability to improve performance over time through software updates.

With a height of 68.3 cm stowed / 160 cm raised, the R1-A7-D is designed to operate at a mid-range level — suitable for navigating under tables, around furniture, and through standard doorways without issue. This compact-but-capable size balances visibility with maneuverability.

The R1-A7-D weighs ~32 kg. Weight affects stability, portability, floor compatibility, and how the robot interacts with its environment.

The R1-A7-D offers Approx. 1.5 hours (battery-powered; external power also supported) of battery life per charge. Battery life is one of the most critical real-world performance metrics for any mobile robot. It determines how much work the robot can accomplish in a single session before needing to recharge. For humanoid robots, this runtime should be evaluated against the size of the area you need covered and the intensity of the tasks involved. Robots with self-charging capability can partially compensate for shorter battery life by autonomously returning to their dock.

The R1-A7-D runs on 8-core body processor plus 8-core head processor with 10 TOPS head-module compute; optional NVIDIA Jetson Orin 40–100 TOPS module for its artificial intelligence capabilities. The AI platform determines how intelligently the robot behaves — from basic reactive responses to sophisticated scene understanding, natural language processing, and adaptive learning. A more advanced AI platform generally means better obstacle avoidance, more natural interaction, and the ability to improve performance over time through software updates.

At 182cm, the H2 Plus is roughly the height of an average adult human, which allows it to interact naturally with human-designed environments including countertops, doorways, and shelving at standard heights. This size is important for robots that need to work alongside people in factories, warehouses, or homes.

The H2 Plus weighs About 70kg with battery. Weight affects stability, portability, floor compatibility, and how the robot interacts with its environment.

The H2 Plus offers About 3 hours of battery life per charge. Battery life is one of the most critical real-world performance metrics for any mobile robot. It determines how much work the robot can accomplish in a single session before needing to recharge. For research robots, this runtime should be evaluated against the size of the area you need covered and the intensity of the tasks involved. Robots with self-charging capability can partially compensate for shorter battery life by autonomously returning to their dock.

The H2 Plus runs on NVIDIA Jetson AGX Thor T5000 with Blackwell GPU, 2,070 FP4 TFLOPS, 14-core Arm CPU, and 128GB unified memory for its artificial intelligence capabilities. The AI platform determines how intelligently the robot behaves — from basic reactive responses to sophisticated scene understanding, natural language processing, and adaptive learning. A more advanced AI platform generally means better obstacle avoidance, more natural interaction, and the ability to improve performance over time through software updates.