Product Selection Guide | How to Choose Among miniHexa, SpiderPi, JetHexa, and ROSpider Hexapod Robots?

Hexapod robots, with their multi-joint biomimetic structure, offer flexible posture control and serve as excellent platforms for learning robot motion control, environmental perception, and intelligent decision-making.

Many hexapod platforms exist on the market, each targeting different hardware levels and learning directions. Some focus on basic motion control for embedded beginners, while others go deeper into machine vision, ROS 2, and embodied intelligence. Matching the platform to your technical background and application needs is essential.

Hiwonder offers four hexapod platforms—miniHexa, SpiderPi, JetHexa, and ROSpider—to meet different learning needs. All use an 18-DOF motion chassis that supports 360° omnidirectional movement and continuous adjustment. Combined with proprietary inverse kinematics algorithms, they allow progressive learning of hexapod biomimetic motion control. Their application depth and feature configurations differ:

miniHexa: Compact Desktop Hexapod Entry Platform

Positioning: Lightweight design focused on embedded development and basic hexapod motion control.

Hardware: Based on an ESP32 controller. Uses HPD-0127 micro dual-axis high-speed servos with 1.8 kg·cm stall torque. Supports optional ESP32-S3 vision module, AI voice interaction module, or WonderLLM AI large model module. Can also expand with touch sensors, infrared obstacle avoidance sensors, and dot-matrix modules.

Battery: 2200 mAh lithium battery, approximately 60 minutes of runtime.

Core advantages: Beginner-friendly with native support for Scratch, Arduino, and Python. Covers the full path from embedded entry to intermediate development. Helps build a solid understanding of hexapod motion control, sensor applications, and basic AI vision. The AI large model module also enables an initial experience with multimodal AI interaction.

SpiderPi: Advanced Hexapod for AI Vision Learning

Positioning: Focuses on OpenCV image processing while covering both robot motion control and AI vision application learning.

Hardware: Based on Raspberry Pi 4B/5. Uses LX-224HV high-voltage bus servos with up to 20 kg·cm torque. Equipped with a 2-DOF vision pan-tilt (170° wide-angle, 300,000 pixels) and a glowing ultrasonic sensor. Supports AI voice interaction module expansion. The official SpiderPi Pro version adds a vision robotic arm for visual grasping and handling practice.

Battery: 2500 mAh lithium battery, approximately 40 minutes of runtime.

Core advantages: Deep integration with the OpenCV image processing library. The flexible head recognition system efficiently supports object tracking, visual line following, tag recognition, and intelligent handling.

JetHexa: ROS 2 Embodied AI Hexapod Development Platform

Positioning: Designed for advanced ROS 2 education and embodied intelligence, focusing on hexapod motion control, 3D vision, and SLAM navigation.

Hardware: Optional Raspberry Pi 5 or NVIDIA Jetson controller. Joints use HX-35H bus servos with up to 35 kg·cm torque. Includes an LCD screen. Supports optional 2-DOF camera pan-tilt, Orbbec Aurora930 Pro depth camera, AI voice interaction box, and LiDAR options such as Guoke COIN-D6 (4000 samples/s) or Slamtec C1 (5000 samples/s).

Battery: 3500 mAh lithium battery, approximately 60 minutes of runtime.

Core advantages: Excellent hardware configuration and strong computing power. Supports high-precision 3D point clouds and three-dimensional environmental perception, ideal for SLAM navigation and high-precision hexapod control learning. Deep deployment of AI large models and the OpenClaw agent meets the needs of advanced embodied AI development.

ROSpider: ROS 2 Embodied AI Hexapod Mobile Manipulation Platform

Positioning: Focuses on full-stack ROS 2 development and explores combined hexapod motion with robotic arm operation.

Hardware: Optional Raspberry Pi 5 or NVIDIA Jetson controller. Joints use HX-35H bus servos with up to 35 kg·cm torque, plus a 6-DOF vision robotic arm. Supports expansion with LiDAR (STL-19P D500, 5000 Hz scan frequency), Aurora930 Pro depth camera, AI voice interaction box, or integrated 6-microphone array.

Battery: 6000 mAh lithium battery, approximately 60 minutes of runtime.

Core advantages: Combines all-terrain biomimetic structure with mobile manipulation. Offers excellent obstacle-crossing and 360° omnidirectional movement capabilities. With the 3D vision robotic arm, multimodal AI large models, and embodied AI framework, it enables autonomous navigation, precise object grasping, and intelligent sorting in complex spaces.

Controller Platform

ESP32: Microcontroller architecture suitable for lightweight embedded control, sensor data collection, and low-level logic.

Raspberry Pi: General-purpose Linux computing platform with a mature ecosystem. Suitable for Python programming, OpenCV image processing, and basic ROS 2 applications.

NVIDIA Jetson: Provides strong on-device GPU computing power. Supports local deep learning deployment, AI large model operation, and complex algorithm processing.

Joint Servos

Although all four robots have an 18-DOF motion chassis, servo strength increases progressively: miniHexa uses 1.8 kg·cm micro servos, SpiderPi uses 20 kg·cm high-voltage bus servos, and JetHexa and ROSpider use 35 kg·cm high-torque bus servos. Load capacity, burst power, and all-terrain stability improve accordingly.

Sensing Hardware and Operation Capability

miniHexa primarily uses ultrasonic, infrared obstacle avoidance, lightweight vision modules, or AI large model modules, focusing on motion control and basic interaction.

SpiderPi uses a 2-DOF vision pan-tilt for planar visual tracking, line following, and tag recognition.

JetHexa adds TOF LiDAR and a 3D depth camera for three-dimensional point cloud collection and autonomous SLAM mapping and navigation.

ROSpider builds on 3D perception and LiDAR with an additional 6-DOF vision robotic arm, advancing from pure chassis movement to mobile chassis plus precise 3D spatial grasping.

Q1: Which one is best for complete beginners?

miniHexa is the best starting point. It supports Scratch graphical block programming so you can control gaits and sensors without writing code. After learning the basics, you can move to Arduino and Python. It is ideal as a first hexapod.

Q2: Both SpiderPi and JetHexa can do visual handling. What is the essential difference?

SpiderPi’s visual handling relies on 2D camera recognition and basic gait adjustment, suitable for learning image processing and control logic. JetHexa includes a 3D depth camera and LiDAR, enabling complex 3D environment reconstruction, SLAM mapping and navigation, and large-model intelligent path planning under the ROS 2 framework—a higher level of embodied AI development.

Q3: How do the different LiDARs differ in SLAM mapping and obstacle avoidance navigation?

COIN-D6 samples at 4,000 times per second and provides balanced performance for basic SLAM, path planning, and autonomous avoidance teaching. Slamtec C1 and STL-19P D500 sample at 5,000 times per second and offer better noise resistance, point cloud detail, and performance with algorithms such as Gmapping, Hector, and Cartographer in larger scenes.

Q4: How does the AI large model experience differ between miniHexa and JetHexa?

miniHexa uses optional modules such as WonderLLM for lightweight multimodal voice or vision interaction. JetHexa and ROSpider, as advanced embodied AI platforms, have stronger on-device computing power and deploy AI large models plus the OpenClaw agent. They better support multimodal instruction parsing, complex environment reasoning, and full-stack embodied AI algorithm validation.

Q5: I am a university or research user working on embodied AI and academic experiments. Should I choose JetHexa or ROSpider?

If your focus is on pure chassis movement, SLAM 3D navigation, multimodal large models, and algorithm validation, JetHexa provides sufficient computing power and environmental perception. If your research involves spatial interaction, intelligent sorting, or grasping and handling, choose ROSpider with its 6-DOF vision robotic arm.