Facts

• Company: Unitree
• Product type: Quadruped
• Status: commercial
• Website: https://www.unitree.com/go2

Profile sections

• Summary
• Overview
• Market context
• Audience
• Workflow
• Capability
• Deployment environment
• Positioning tags
• Specifications
• Tags
• Timeline / Research
• Comparable Systems
• Sources

Summary

Go2 is Unitree compact quadruped robot for education, research and mobile robotics development. It combines legged mobility, sensing and app-based control for users testing navigation, perception and robot-dog applications.

Overview

Go2 is Unitree compact quadruped robot for education, research and mobile robotics development. It combines legged mobility, sensing and app-based control for users testing navigation, perception and robot-dog applications.

Market context

Compact quadruped robot with 4D LiDAR and 360° perception.

• Target environment: Research labs, university robotics programs and light-duty inspection environments.
• Workflow context: Inspection, research, development and light operational tasks using onboard perception and autonomous navigation.
• Customer context: Robotics researchers, developers and inspection teams seeking an accessible quadruped platform.
• Deployment model: Commercial product sold as a development and light-operations platform.
• Commercial maturity: Commercially available quadruped with broad developer adoption due to accessible pricing.
• Adoption constraints: Compact form factor limits payload and terrain capability compared to industrial-grade quadrupeds.
• Market position: Unitree's accessible quadruped platform, widely used for research and light operational use. Sits below the heavier-duty B2.
• Adjacent products: B2, Spot, ANYmal

Audience

Developers

Workflow

Inspection, research, development and light operational tasks using onboard perception and autonomous navigation.

Capability

Autonomous navigation

Deployment environment

Research labs, university robotics programs and light-duty inspection environments.

Positioning tags

• Audience: Developers
• Workflow: Inspection, research, development and light operational tasks using onboard perception and autonomous navigation.
• Capability: Autonomous navigation
• Deployment environment: Research labs, university robotics programs and light-duty inspection environments.

Specifications

• Dimensions: Standing dimensions 70 x 31 x 40 cm
• Payload Capacity: Approx. 7 to 8 kg nominal payload depending on variant; max approx. 10 to 12 kg depending on variant
• Speed: 0 to 2.5 m/s, 3.5 m/s, or 3.7 m/s depending on variant; max approx. 5 m/s on higher variants
• Voltage: 28 V to 33.6 V
• Max Power: About 3000 W
• Obstacle Clearance: Max climb/drop height about 15 to 16 cm depending on variant
• Slope: Max climb angle 30° or 40° depending on variant
• Perception System: Unitree 4D LiDAR L2 with 360° x 96° hemispherical recognition
• Task Scope: Consumer, developer, and mobile robotics applications

Tags

• Peer Group: Industrial inspection quadruped
• Workflow: Industrial inspection
• Capability: Autonomous navigation
• Deployment Environment: Research lab
• Stack Layer: Finished robot platform
• Commercial Stage: Developer platform

Timeline / Research

• 2026-06-24 - Research Publication - RoboAtlas paper evaluates contextual active SLAM on Unitree Go2 - RoboAtlas combines frontier exploration, global semantic-map reasoning, and egocentric VLM reasoning for contextual A…
• 2026-06-24 - Research Publication - StairMaster paper deploys hollow-stair locomotion policy on Unitree Go2 - StairMaster is a three-stage reinforcement-learning framework for quadruped locomotion on risky hollow stairs.
• 2026-06-23 - Research Publication - fARfetch: Enabling Collocated AR-HRC in Large Visually Diverse Environments with VLM-Driven AR Content Adaptation - The paper uses Unitree Go2 for collocated augmented-reality human-r…
• 2026-06-18 - Research Publication - ForEnt dataset paper collects forest-entrapment data with Unitree Go2 - ForEnt is a multi-modal dataset for studying quadruped robot entrapments in forest environments.
• 2026-06-15 - Research Publication - SemGeoNav paper evaluates safety-guided visual navigation on Unitree Go2 - SemGeoNav integrates semantic reasoning with geometric planning for safer visual navigation in open environm…
• 2026-06-14 - Research Publication - FlashNav paper trains deployable navigation policies for Unitree Go2 within seconds - FlashNav is a GPU-first framework for ultra-fast range-based robot navigation policy training.
• 2026-06-14 - Research Publication - LoComposition paper deploys terrain-adaptive quadruped locomotion on Unitree Go2 - LoComposition decomposes quadruped locomotion into task rewards, operational constraints, energy min…
• 2026-06-12 - Research Publication - Kine2Go paper releases kinematic dataset for Unitree Go2 gaits and motions - Kine2Go introduces a kinematic dataset for the Unitree Go2 robot covering diverse gaits and motions.
• 2026-06-09 - Research Publication - AgniNav paper transfers monocular local navigation to Unitree Go2 - AgniNav standardizes cross-embodiment monocular local navigation through a configurable collision-envelope represen…
• 2026-06-06 - Research Publication - Semantic-skill discovery paper validates multi-quadruped coordination on Unitree Go2 teams - The paper studies continual quadruped robot coordination through semantic skill discovery.
• 2026-06-02 - Research Publication - Library service-robot navigation paper builds ROS 2 stack on Unitree Go2 Edu - The paper presents a ROS 2 autonomous-navigation system for a Unitree Go2 Edu quadruped operating in lib…
• 2026-03-30 - Research Publication - DFKI publishes Unitree Go2 quadruped state-estimation dataset - DFKI Robotics Innovation Center published a Unitree Go2 dataset for legged-robot state estimation.
• 2026-03-02 - Research Publication - Safe whole-body loco-manipulation paper validates model-learning control on Unitree Go2 - The paper combines model-based admittance control for manipulation with a reinforcement-learn…
• 2026-02-11 - Research Publication - External stair-climbing paper trains and transfers policies on Unitree Go2 - The paper studies a two-stage reinforcement-learning approach for improving quadruped performance on U-sha…
• 2026-02-07 - Research Publication - UEREBot paper deploys safe obstacle-evasion locomotion on Unitree Go2 - The UEREBot paper proposes a hierarchical safe locomotion framework for unstructured terrain and high-speed dyn…
• 2025-10-16 - Research Publication - Risk-aware reinforcement-learning paper evaluates adaptive quadruped locomotion on Unitree Go2 - The paper trains risk-conditioned quadruped locomotion policies using CVaR-constrained…
• 2025-07-27 - Research Publication - Bipedalism paper deploys risk-adaptive loco-manipulation policy on Unitree Go2 - The paper proposes a risk-adaptive reinforcement-learning framework that enables a quadruped to walk o…
• 2023-07-12 - Product Launch - Unitree launches Go2 embodied-AI quadruped robot - Unitree unveiled Go2 as a new-generation consumer-grade quadruped robot with 4D LiDAR and embodied-AI positioning.

Comparable Systems

• Lite3 - Deep Robotics - Quadruped
• Lynx M20S - Deep Robotics - Quadruped
• X30 - Deep Robotics - Quadruped
• Spot - Boston Dynamics - Quadruped
• B2 - Unitree - Quadruped
• AS2 - Unitree - Quadruped
• MagicDog-W - MagicLab - Quadruped
• A2 - Unitree - Quadruped