Description

Seeed-100064063 is the Raspberry Pi AI HAT+ 13 TOPS, a Hailo-8L AI accelerator board designed for Raspberry Pi 5 edge AI deployment. It adds a dedicated Hailo-8L NPU with 13 TOPS INT8 compute through the Raspberry Pi 5 PCIe 3.0 interface, enabling low-latency local inference for computer vision, object detection, image segmentation, face recognition and other AI workloads without depending on cloud processing.

The Raspberry Pi AI HAT+ is useful when a Raspberry Pi 5 project needs real-time AI acceleration but must remain compact, cost-efficient and power-efficient. It works with mainstream AI frameworks and the Raspberry Pi software ecosystem, including rpicam-apps integration for camera-based object detection and face recognition. The HAT+ mechanical design keeps the system compatible with the official Raspberry Pi Active Cooler, helping maintain stable operation during heavier inference loads.

For developers searching for a Raspberry Pi 5 AI accelerator, Hailo-8L module, Raspberry Pi AI HAT, 13 TOPS NPU board, edge AI accelerator, computer vision HAT or local inference hardware, Seeed-100064063 provides a practical path from prototype to deployed smart camera, industrial inspection, robotics or home automation system.

Features

• Onboard Hailo-8L AI accelerator with 13 TOPS INT8 performance for high-performance Raspberry Pi 5 AI tasks.
• PCIe 3.0 connection to the Raspberry Pi 5 for low-latency data transfer and efficient edge inference.
• Accelerates object detection, image segmentation, face recognition, package detection, people counting, license plate recognition and robotic navigation.
• Compatible with mainstream AI frameworks and suitable for computer vision models running locally at the edge.
• Raspberry Pi OS friendly: latest Raspberry Pi OS can recognize supported drivers without manual kernel driver compilation.
• rpicam-apps integration supports Hailo post-processing for camera AI workflows from command-line tools.
• HAT+ mechanical standard with ventilation cutouts for compact Raspberry Pi 5 installations.
• Compatible with the official Raspberry Pi Active Cooler; the AI HAT+ sits above the cooler for thermal stability.
• Power-efficient AI inference for always-on monitoring nodes, mobile robots, smart security and battery-aware edge projects.
• Includes GPIO stacking header and screw set for a practical Raspberry Pi 5 AI HAT installation.

Specifications

Applications

• Smart home security with local human, pet, vehicle and package detection to reduce false alarms.
• Industrial automation for part sorting, defect inspection, surface scratch detection and production-line visual inspection.
• Smart traffic and retail analytics including people counting, heatmap generation and license plate recognition for small parking lots.
• Robotics development with real-time scene parsing, obstacle detection and camera-based navigation.
• Education and maker projects such as gesture control, computer vision experiments and Raspberry Pi 5 AI model demos.

Product List

• 1 x Raspberry Pi AI HAT+ 13 TOPS
• 1 x 16mm GPIO stacking header
• 1 x Screw set

Other Content

The official hardware diagrams help clarify how the Hailo-8L AI accelerator integrates with the Raspberry Pi 5 platform and why the HAT+ mechanical design matters for thermal stability.

Hardware overview. This diagram identifies the AI HAT+ layout and shows the Hailo-8L accelerator board as a compact add-on for Raspberry Pi 5 edge AI projects.Thermal and mechanical stack. The AI HAT+ is designed to sit above the official Raspberry Pi Active Cooler, keeping airflow and mechanical clearance in mind for sustained inference workloads.

Resources

• Seeed product PDF for Raspberry Pi AI HAT+ 13 TOPS
• Raspberry Pi Product Information Portal

FAQ

Does this board work with Raspberry Pi 5?

Yes. The Raspberry Pi AI HAT+ is designed for Raspberry Pi 5 and connects through the PCIe 3.0 interface.

What does 13 TOPS mean?

13 TOPS refers to the Hailo-8L accelerator performance for INT8 AI inference, useful for real-time edge AI workloads such as object detection and image segmentation.

Can it run AI locally without cloud processing?

Yes. The purpose of this Hailo-8L AI accelerator is to run local inference on Raspberry Pi 5, reducing latency and cloud dependency.

Is it compatible with the official Raspberry Pi Active Cooler?

Yes. The HAT+ design sits above the official Active Cooler for better thermal stability under AI inference loads.