Comparison table
 Bittle XPetoi |  MentorPi M1Hiwonder |  ROSMASTER M1Yahboom |  XGO-Rider2Luwu Dynamics |  UGV RoverWaveshare |  Mini Pupper 2MangDang |  PuppyPi ProHiwonder | |
|---|---|---|---|---|---|---|---|
| Price | |||||||
| Price range | $289 – $469 | $430 – $880 | $349.90 – $1,169.90 | $549 | $245 – $727 | $649 | $590 – $1,180 |
| Version compared | Bittle X $319 plus the $39 Petoi AI Vision Camera | Starter with Pi 5 2 GB $429.99 | Standard with RDK X5 $529.90 | XGO-Rider2 $549 from Luwu | PT ROS 2 kit with Pi 5 4 GB $726.99 | Pre-assembled $649, plus the $119 lidar and $8 camera, about $776 (est.) | Pro Kit with Pi 5 2 GB $839.99 |
| Also needed | Nothing listed | Nothing listed | Nothing listed | Nothing listed | Three 18650 cells | The camera module, sold out in October 2026 | Nothing listed |
| Spotting the cat | |||||||
| On-board computer | ESP32 controller, with detection on the camera module | Raspberry Pi 5 2 GB | RDK X5 | Raspberry Pi CM5 | Raspberry Pi 5 4 GB | Raspberry Pi CM4 | Raspberry Pi 5 2 GB |
| AI accelerator | On the camera module, which runs YOLOv5 and YOLOv8 | None | Up to 10 TOPS on the RDK X5 | None | On the OAK-D Lite depth camera, with no official detection example | None | None |
| Cat detection from the maker | 80-class detector that likely includes cat, with no cat behavior (unconfirmed) | YOLO lessons with custom training, no ready cat model found | YOLOv11 course for the RDK X5 (cat model unconfirmed) | Face tracking, with object detection unconfirmed | Official MobileNet-SSD example whose 20 classes include cat | OAK-D Lite object following demo for ROS 1 only (unconfirmed) | Color and tag tracking and MediaPipe faces, no YOLO lessons found |
| Camera | Petoi AI Vision Camera | 640 × 480 USB camera, 170° view | 2 MP on a pan-tilt | 5 MP OV5647 | 5 MP 160° on a pan-tilt, plus OAK-D Lite depth | OV5647 module, sold separately | 480p 130° on a 2-axis mount |
| Searching other rooms | |||||||
| Lidar | None | COIN-D6 ToF lidar | T-MINI Plus, 12 m range | None | D500 360° | STL-06P, $119 extra | ToF lidar |
| Mapping and navigation | No, the camera module cannot run SLAM, so it needs an added Raspberry Pi | ROS 2 slam_toolbox mapping and navigation lessons | ROS 2 lidar mapping and navigation course | No lidar or mapping | ROS 2 SLAM and Nav2, plus Gazebo simulation | ROS 2 Humble SLAM and Nav2 documented | ROS 2 lidar mapping and navigation chapters (Pi 5 only) |
| Moving | |||||||
| Drive | 4 legs | Mecanum wheels | Mecanum wheels | 2 wheels on balancing legs | 6 rubber wheels, skid steer | 4 legs | 4 legs |
| Speed | ~0.4 m/s | Not published | Not published | > 1 m/s (maker claim) | 1.3 m/s | ~0.3–0.5 m/s (est.) | ~0.2 m/s (est.) |
| Battery life | Not published | Not published (7.4 V 2200 mAh) | Not published (6,000 mAh) | About 1 h | About 1.5 h | About 30 min | Not published (7.4 V 2200 mAh) |
| Programming | |||||||
| Languages and tools | Python, Arduino, C++ | ROS 2, Python and C++, phone app | ROS 2, Python, OpenCV, MediaPipe, YOLOv11 | Python (xgolib), C++, blocks | Python, OpenCV and MediaPipe, ROS 2 Humble | ROS 2, Python, open source | Python, OpenCV, ROS 1 and ROS 2 |
| LLM support | Official LLM tools | Official LLM tools | Official LLM tools | LLM plans actions | Official LLM tools | Official LLM tools | Official LLM tools |
| Purchasing | |||||||
| Suggested ages | 14+ | 14+ | 14+ | 14+ | 14+ | 14+ | 14+ |
| Shipping to the US | 2–5 days | Next day (Amazon) | ~1–2 wks (est.) | 3–6 wks | ~1–2 weeks direct | ~2–3 weeks | 2–3 days |
Highlighted cells are the strongest in their row. Speeds marked (est.) are not published by the maker. They come from owner measurements or the makers' gait code, explained on each robot page. Prices are US list prices as of September 2026. Each robot's page has full specs, sources and photos.
Which one fits
- UGV Rover is the strongest all-round pick. It is the fastest here at 1.3 m/s, its six rubber wheels suit carpet and thresholds, and it has lidar, an official ROS 2 SLAM and Nav2 setup, and an official on-board detector whose classes include cat. It ships assembled from China, and the three 18650 cells cost extra.
- ROSMASTER M1 with the RDK X5 board is the cheapest way to get both a lidar and a real AI accelerator, at $529.90, with a YOLOv11 course for the board. It comes as a kit, and its Mecanum wheels grip carpet less well than tires.
- MentorPi M1 Starter is the cheapest kit here with a Raspberry Pi 5, lidar and camera, at $429.99, and it ships next day from Amazon. The 2 GB Pi is tight for running ROS 2 navigation and a detector together, and there is no ready-made cat model.
- XGO-Rider2 balances on two wheeled legs and is claimed to pass 1 m/s, with a Raspberry Pi CM5 and a 5 MP camera. It has no lidar or mapping, so it can follow a cat in view but cannot search other rooms without camera-only navigation code.
- Mini Pupper 2 is the legged option with documented ROS 2 SLAM and Nav2. The lidar and camera are extra, which brings it to about $776 (est.), and the battery lasts about 30 minutes.
- PuppyPi Pro adds lidar mapping and navigation to a Raspberry Pi 5 robot dog with Hiwonder's lessons, from $839.99. At about 0.2 m/s (est.) it is the slowest here, and its vision lessons cover colors, tags and faces rather than YOLO.
- Bittle X is the cheapest at about $358 with Petoi's $39 AI camera, which runs YOLO models on the module itself. It can follow a cat in view, but it cannot map or search rooms without an added Raspberry Pi.