AgiBot X2

AgiBot X2

Fully Intelligent Agile Humanoid Robot

Figure 1 — AgiBot X2 in motion

Overview

The AgiBot X2 is a friendly, bipedal humanoid robot built around three core principles: Affable, Agile, and Intelligent. It walks, dances, navigates and interacts with people in ways that feel natural, warm and alive. With 25–30 active degrees of freedom, a high-precision sensor matrix and on-board autonomous decision-making, X2 is equally at home greeting visitors, guiding people through a space, or demonstrating human-like motion at events and exhibitions.

Two configurations share the same platform — the X2 Youth and the X2 Flagship. The Flagship adds depth perception, a 3D LiDAR, a high-compute NVIDIA Orin NX module, and secondary development support.

Three Qualities That Define X2

Affable — Human-like Interaction

Figure 2 — X2 engaging in an interactive task

X2 combines visual, voice, tactile and on-screen facial interaction. It reads user gestures, scene context and conversational context, and responds with natural, expressive body language. A microphone array, mini wireless microphone and dual-channel speaker support clear voice dialogue, while an interactive screen and light effects convey mood and status.

· Visual, voice, touch and facial interaction

· Gesture recognition and context-aware responses

· Expressive screen face and ambient light effects

Agile — Lifelike Motion

Figure 3 — X2 in a relaxed, natural pose

With 25–30 body degrees of freedom — including a 3-DOF waist and a 1-DOF neck — X2 reproduces human-like gait, complex choreography and turns its head toward the sound. The waist twists, bends sideways and leans, and the natural sway of walking mimics the rhythm of a human step.

· Human-like walking gait with natural body sway

· Complex, expressive dance and movement routines

· Head follows voice; playful head-pat touch feedback

Intelligent — Autonomous Tasks

Figure 4 — X2 autonomously navigating an office environment

No remote control required. The Flagship model is equipped with a multi-line 3D LiDAR and an RGB-D camera, enabling it to perceive its surroundings, plan its own walking route and actively avoid obstacles. Decision-making happens on board, so X2 operates independently in real environments.

· Autonomous navigation and route planning

· Active obstacle avoidance via LiDAR and RGB-D vision

· On-board algorithmic decision-making, no RC needed

High-Precision Sensor Matrix

 X2 sees in depth, recognizes faces and gestures, and maps the world around it in three dimensions.

· Interaction RGB camera — face recognition, object detection, gesture recognition, environmental semantic perception

· RGB-D camera (Flagship) — true 3D spatial perception and depth recognition to support manipulation decisions

· Multi-line 3D LiDAR (Flagship) — point-cloud acquisition for high-precision 3D SLAM localization and navigation

· Head touch sensor — capacitive touch on top of the head triggers a playful nodding reaction

Technical Specifications

Note: End effectors (OmniHand 10-DOF dexterous hand, OmniPicker adaptive gripper) are optional, Flagship-only and sold separately. Specifications are subject to the final delivered unit.

What's in the Box

· X2 robot unit x1

· Charger x1

· Handheld remote controller x1

· Mini wireless microphone x1

· Quick start guide x1

· Battery pack x1 + certificate of conformity