Raspberry Pi AI HAT+ 2
A powerful and versatile machine learning HAT for the Pi 5 - run LLMs, VLMs and conventional vision models on low-power, edge hardware.
In stock, ships same business day if ordered before 2PM
Fastest delivery: Tomorrow*
Disclaimer:
For next-day delivery, the shipping address must
be in the AusPost next-day network, eParcel Express must be selected, and the order must be placed
before 2PM AEST Mon-Thurs excluding NSW Public Holidays. Orders may be delayed due to AusPost
pickup timings and order verifications. eParcel Express is typically a 1-day service within the
AusPost next-day network, though it is sometimes 2+ days.
Recommended Essentials:
Building on the success of the original AI HAT, the AI HAT+ 2 delivers next‑generation artificial‑intelligence performance to the Raspberry Pi 5. Powered by the Hailo 10H neural‑network accelerator and equipped with 8 GB of high‑speed LPDDR4X‑4267 SDRAM, this board brings powerful LLM and VLM capabilities, along with the previous models' machine vision capabilities to the edge - perfect for offline, local AI workloads or robotics projects.
Key Features
- Hailo 10H accelerator – up to 40 TOPS @ INT‑4 and 20 TOPS @ INT‑8
- On‑board 8 GB LPDDR4X‑4267 SDRAM for loading larger models directly on the HAT
- Runs large language models (LLMs) such as QWEN and Llama 3 B, and vision models (YOLO, etc.) with the same performance as the 26 TOPS HAT+
- Optimised for power‑efficient edge AI – ideal for local, offline inference
- Supports both LLMs and VLMs, enabling chat‑GPT‑style interactions, home‑automation decision‑making, robotics, and more
- Future‑proof: updates to models will keep the HAT useful into the future
- Uses the PCIe connector – only compatible with Raspberry Pi 5
Compatibility & Installation
- Works with an active cooler mounted to the Pi 5 SBC
- Designed to fit the Official Raspberry Pi 5 Case (Remove the in-built fan and top plate)
- PCIe connection cable included for a clean, low‑profile installation
Highlights
The AI HAT+ 2 delivers up to 40 TOPS (INT‑4) and 20 TOPS (INT‑8), matching the vision‑model performance of the 26 TOPS HAT+ while adding LLM capability. This makes it an ideal platform for:
- Running sophisticated language models for chat‑bot or decision‑support applications
- Processing visual data with YOLO‑style detection and other computer‑vision pipelines
- Deploying AI in resource‑constrained environments where power and space are limited
Product Contents
- AI HAT+ 2 board
- Heatsink for the chipset
- Stacking header
- PCIe connection cable
- Standoffs for secure mounting
Technical Specifications
- Neural‑network accelerator: Hailo 10H
- Memory: 8 GB LPDDR4X‑4267 SDRAM
- Interface: PCIe (Raspberry Pi 5 only)
- Performance: 40 TOPS @ INT‑4, 20 TOPS @ INT‑8
- Power efficiency: Optimised for low‑power edge operation
Resources
- Getting Started Guide
- Blog Post
Building on the success of the original AI HAT, the AI HAT+ 2 delivers next‑generation artificial‑intelligence performance to the Raspberry Pi 5. Powered by the Hailo 10H neural‑network accelerator and equipped with 8 GB of high‑speed LPDDR4X‑4267 SDRAM, this board brings powerful LLM and VLM capabilities, along with the previous models' machine vision capabilities to the edge - perfect for offline, local AI workloads or robotics projects.
Key Features
- Hailo 10H accelerator – up to 40 TOPS @ INT‑4 and 20 TOPS @ INT‑8
- On‑board 8 GB LPDDR4X‑4267 SDRAM for loading larger models directly on the HAT
- Runs large language models (LLMs) such as QWEN and Llama 3 B, and vision models (YOLO, etc.) with the same performance as the 26 TOPS HAT+
- Optimised for power‑efficient edge AI – ideal for local, offline inference
- Supports both LLMs and VLMs, enabling chat‑GPT‑style interactions, home‑automation decision‑making, robotics, and more
- Future‑proof: updates to models will keep the HAT useful into the future
- Uses the PCIe connector – only compatible with Raspberry Pi 5
Compatibility & Installation
- Works with an active cooler mounted to the Pi 5 SBC
- Designed to fit the official Raspberry Pi 5 case (confirmation pending); if not, a compatible case such as the KKSB HAT Case is recommended
- PCIe connection cable included for a clean, low‑profile installation
Highlights
The AI HAT+ 2 delivers up to 40 TOPS (INT‑4) and 20 TOPS (INT‑8), matching the vision‑model performance of the 26 TOPS HAT+ while adding LLM capability. This makes it an ideal platform for:
- Running sophisticated language models for chat‑bot or decision‑support applications
- Processing visual data with YOLO‑style detection and other computer‑vision pipelines
- Deploying AI in resource‑constrained environments where power and space are limited
Product Contents
- AI HAT+ 2 board
- Heatsink for the chipset
- Stacking header
- PCIe connection cable
- Standoffs for secure mounting
Technical Specifications
- Neural‑network accelerator: Hailo 10H
- Memory: 8 GB LPDDR4X‑4267 SDRAM
- Interface: PCIe (Raspberry Pi 5 only)
- Performance: 40 TOPS @ INT‑4, 20 TOPS @ INT‑8
- Power efficiency: Optimised for low‑power edge operation
- Works with Active cooler
WE CAN TODO CONFIRM
- Fits in Official case
- If not, what case? - KKSB HAT CASE?
Discussion points
- Building on success of original AI HAT
- Capable of running LLMs and VLMs - along with the same vision models (YOLO etc) with the same level of performance as the 26TOPS HAT+
- Made for AI on the edge, or local and offline
- Power efficient
- Expected to run models like QWEN and Llama (3B)
- A lot of the questions we got were 'Can I run something similar to chatgpt on the AI HAT" - now you can
- Would be perfect for general questions, and decision making - home automation, robots,
- As models get better so will this HAT
- Uses the PCIe connector - only compatible with the Pi 5
Short desc options
- Give your Pi 5 an AI boost: low‑power, offline inferencing that can power chat assistants, vision detection and decoding, and robots.
- A powerful and versatile machine learning HAT for the Pi 5 - run LLMs, VLMs and conventional vision models.
- Transform your Pi 5 into an edge‑AI powerhouse, with low‑power offline operation for chat‑bots, visual detection, and robotics.
This product is listed in:
Raspberry Pi>Model B Hats & PlatesHome Lab>Machine Learning
Raspberry Pi>Raspberry Pi 5>Raspberry Pi 5 Accessories
Exact shipping can be calculated on the view cart page (no login required).
Products that weigh more than 0.5 KG may cost more than what's shown (for example, test equipment, machines, >500mL liquids, etc).
We deliver Australia-wide with these options (depends on the final destination - you can get a quote on the view cart page):
- $3+ for Stamped Mail (typically 10+ business days, not tracked, only available on selected small items)
- $7+ for Standard Post (typically 6+ business days, tracked)
- $11+ for Express Post (typically 2+ business days, tracked)
- Pickup - Free! Only available to customers who live in the Newcastle region (must order online and only pickup after we email to notify you the order is ready). Orders placed after 2PM may not be ready until the following business day.
Non-metro addresses in WA, NT, SA & TAS can take 2+ days in addition to the above information.
Some batteries (such as LiPo) can't be shipped by Air. During checkout, Express Post and International Methods will not be an option if you have that type of battery in your shopping cart.
International Orders - the following rates are for New Zealand and will vary for other countries:
- $12+ for Pack and Track (3+ days, tracked)
- $16+ for Express International (2-5 days, tracked)
If you order lots of gear, the postage amount will increase based on the weight of your order.
Our physical address (here's a PDF which includes other key business details):
40 Aruma Place
Cardiff
NSW, 2285
Australia
Take a look at our customer service page if you have other questions such as "do we do purchase orders" (yes!) or "are prices GST inclusive" (yes they are!). We're here to help - get in touch with us to talk shop.
Have a product question? We're here to help!
Videos
View AllGuides
YOLO Pose Estimation on the Raspberry Pi AI Hat+ | Writing Custom Python Code
YOLO Object Detection on the Raspberry Pi AI Hat+ | How to Write Custom Python Code
How to Add an SSD To Your Raspberry Pi 5 with the M.2 HAT
Raspberry Pi 5 Vs Raspberry Pi 4 Model B | Comparison & Benchmarking
Projects
safePower
ESP32 IR-Triggered Camera with Home Assistant Integration
LuMa 1 - The Neo Disco Ball
Educational Workshops
Raspberry Pi Workshop for Beginners
Makers love reviews as much as you do, please follow this link to review the products you have purchased.
Product Comments