The arrival of the Raspberry Pi AI HAT+ 2 represents a new step forward for those who want to work with artificial intelligence directly on a Raspberry Pi 5 without relying on the cloud. This expansion board combines a dedicated neural accelerator and its own memory to offload much of the AI ​​work from the main processor and enable more ambitious applications in the field of generative AI.
With a suggested retail price of around $130 (the final price in Spain and the rest of Europe will depend on official distributors), the AI ​​HAT+ 2 is positioned as a relatively affordable option for experimenting with language models and computer vision at the edge. It's not a replacement for high-end solutions, but it is a compelling piece of equipment for IoT, automation, prototyping, and training projects.
What exactly is the Raspberry Pi AI HAT+ 2
The Raspberry Pi AI HAT+ 2 is an expansion board for the Raspberry Pi 5 that connects via the PCI Express interface integrated into the motherboard. It is the direct successor to the first AI HAT+, released in 2024, which focused almost entirely on computer vision tasks thanks to the Hailo-8 and Hailo-8L accelerators.
In this new model, Raspberry Pi Holdings plc has opted for the Hailo-10H neural network accelerator and added 8 GB of dedicated LPDDR4X memory directly on the board. This combination is designed to support generative AI workloads , especially large, small-scale language models and vision-language models, directly on the device.
By having DRAM memory on the HAT itself, the system can separate the resources of the Raspberry Pi 5 (CPU, GPU, and main RAM) from those of the AI ​​accelerator. This allows the main board to handle application logic, communications, and user interface, while the Hailo-10H coprocessor takes over most of the inference.
Hardware, performance and differences with the first AI HAT+
The heart of the new accessory is the Hailo-10H NPU, a neural network accelerator that, according to Raspberry Pi and Hailo, offers up to 40 TOPS of inference performance . The documentation mentions figures for INT4 and INT8 , indicating that some of the peak performance is achieved through aggressive quantization, common in edge AI deployments. However, other embedded platforms like the Nvidia Jetson T5000 offer alternatives with different architectures.
The Hailo-10H's power consumption is limited to around 3W , which helps keep the system within reasonable temperature and power limits—crucial for embedded projects and compact enclosures. However, this restriction means that its raw performance won't always surpass that of the Raspberry Pi 5 working alone on certain tasks, especially when comparing workloads highly optimized for the integrated CPU and GPU.
Compared to the first AI HAT+, the main improvement is the jump from the 13/26 TOPS of the Hailo-8L and Hailo-8 to 40 TOPS, in addition to the inclusion of 8 GB of onboard RAM. The original model was primarily designed for object detection, pose estimation, and segmentation, while this second generation opens the door to LLMs and multimodal models , while maintaining its vision capabilities.
Raspberry Pi has emphasized that integration with the camera environment should be virtually seamless for existing projects, as integration with the company's software stack is maintained, avoiding the need to rewrite all computer vision code from scratch.
Supported AI models and local LLM usage
One of the most striking features is the ability to run language models locally on the Raspberry Pi 5, leveraging the power of the NPU and the dedicated memory of the HAT. During the launch, the company mentioned a number of LLMs with between 1.000 and 1.500 billion parameters as a starting configuration.
Among the models cited are DeepSeek-R1-Distill, Llama 3.2, Qwen2, Qwen2.5-Instruct, and Qwen2.5-Coder . These relatively compact models enable use cases such as basic chat, code generation, text translation, and scene description, all without sending data to external servers.
The board is designed to leverage quantization and optimized inference techniques , enabling models of this size to run with low latency and minimal power consumption. Raspberry Pi and Hailo have demonstrated examples of language translation and responses to simple questions, all executed entirely on the device.
However, it's important to understand that AI HAT+ 2 isn't designed for massive models like the full versions of ChatGPT, Claude, or the largest LLMs from Meta, whose number of parameters ranges in the hundreds of billions or even trillions. In those cases, the barrier isn't just computing power, but primarily the amount of memory required.
Memory limitations and comparison with a 16GB Raspberry Pi 5
While the 8 GB of dedicated LPDDR4X RAM represents a significant improvement over the first-generation AI HAT+, this figure has clear implications. Many medium-sized, context-rich quantized language models can easily require more than 10 GB of memory , so for now, the accessory is geared towards smaller models or those with limited contexts.
When compared directly to a 16GB Raspberry Pi 5, the motherboard with the largest memory configuration still offers more headroom for loading large models entirely into RAM , especially if you forgo some memory for other tasks and dedicate the system almost exclusively to AI.
In practice, combining a Pi 5 and the AI ​​HAT+ 2 makes sense when you want to separate functions : letting the Hailo-10H handle inference while the main board maintains a lightweight desktop, web servers, automation logic, or additional services. This way, the system remains usable while running language or vision models.
For those who only want to chat with a simple local AI, translate texts, or test basic assistants, the AI ​​HAT+ 2's balance of power, consumption, and cost may be sufficient. However, for the memory required by large models or for very extensive contexts, it will still be more practical to use devices with more memory or the cloud.
Computer vision and simultaneous model execution
One of the advantages of this new generation is that it doesn't sacrifice the computer vision capabilities already offered by the first AI HAT+. The Hailo-10H maintains very similar performance to the Hailo-8 when running object detection and tracking models, human posture estimation, or scene segmentation.
Raspberry Pi indicates that the AI ​​HAT+ 2 is capable of simultaneously running vision and language models , making it interesting for projects that combine cameras and text: for example, smart cameras that generate descriptions, surveillance systems that classify events, or devices that offer summaries of what is happening in a scene.
In practical terms, scenarios mentioned include the use of YOLO -type models for real-time object detection, with refresh rates that can reach around 30 frames per second depending on the model's complexity and resolution. The idea is that the HAT handles this processing while the Raspberry Pi 5 manages storage, the interface, and sending alerts.
The software ecosystem is still maturing: although there are examples and official frameworks, the parallel execution of multiple models (vision, language, multimodal) remains an evolving field. In any case, integration with the Raspberry Pi camera stack greatly simplifies the setup process for those already working with official camera modules.
Use cases in Spain and Europe: IoT, industry and prototypes
The combination of low power consumption, small size, and local AI execution aligns well with many of the digitalization trends seen in Spain and other European countries. In industrial environments, where constant cloud access is not always guaranteed or is undesirable due to confidentiality concerns, this type of solution, compared to alternatives like Nvidia chip integration, can be particularly useful.
Among the most frequently mentioned uses in the documentation are industrial automation, process control, and facilities management applications . Visual inspection systems on production lines, real-time anomaly detection, access control, and people counting are some of the applications that could benefit from combining vision and lightweight language models.
In the field of home and commercial IoT, the AI ​​HAT+ 2 can serve as a basis for local assistants, smart panels that interpret sensor data, cameras that describe scenes, or video analytics devices that do not have to upload images to external servers, helping to comply with increasingly demanding privacy regulations in the European Union.
It's also a viable option as a development kit for companies and startups looking to integrate the Hailo-10H chip into their own end products. Testing the solution on a Raspberry Pi 5 allows you to validate performance, power consumption, and stability before embarking on custom hardware designs.
User profile: from makers to professional developers
The AI ​​HAT+ 2 targets several user profiles simultaneously. On one hand, it appeals to the maker and enthusiast community already familiar with Raspberry Pi who want to take their projects further with generative AI or advanced vision without investing in dedicated workstations or ongoing cloud services.
On the other hand, it clearly targets professional developers and startups seeking an affordable testing environment for embedded AI. Compared to solutions with discrete GPUs or NPUs for industrial PCs, this HAT offers a compact form factor, low power consumption, and a more affordable price, at the expense of the performance levels of much more expensive platforms.
In educational and technical training environments in Spain and Europe, it can become a useful tool to bring technical training closer to practice, allowing students to experiment with real models on relatively inexpensive hardware.
However, for users who only want to run language models locally with the maximum possible margin, a Raspberry Pi 5 with 16 GB of RAM without HAT may still be a simpler option, provided it is accepted that the CPU and GPU of the main board will be the ones that handle all the inference.
Software integration and resources to get started
Raspberry Pi has emphasized that the AI ​​HAT+ 2 uses a PCIe interface and is compatible with the familiar Raspberry Pi 5 environment, reducing friction for those already familiar with the ecosystem. Communication with the HAT is via the PCIe interface, and dedicated drivers and libraries allow AI workloads to be routed to the Hailo-10H.
Hailo provides users with a GitHub repository and a Developer Zone offering code examples, pre-built models, tutorials, and frameworks to leverage both generative AI and computer vision. This makes it easy to start prototyping without having to build the entire stack from scratch.
At launch, several ready-to-install language models are already being advertised , with the promise of expanding the catalog with larger variants or those fine-tuned to specific use cases. Furthermore, the possibility of using techniques like LoRA (Low-Rank Adaptation) to adapt the models to specific tasks without having to completely retrain them is mentioned.
As is often the case with these types of solutions, the actual experience will depend on the software's maturity level . Some analysts point out that there is still room for improvement in stability, tools, and support for simultaneous execution of multiple models, but the trend points toward increasingly polished integration within the Raspberry Pi ecosystem.
Price, availability and practical aspects
As for the price, the Raspberry Pi AI HAT+ 2 has been announced with a suggested retail price of $130. In Spain and other European countries, the final price will depend on the exchange rate, taxes, and the policies of each retailer , so slight variations are expected.
The board is compatible with all Raspberry Pi 5 versions, from models with 1GB to 16GB of RAM , and connects via the SBC's own PCIe interface. This eliminates the need for additional adapters and simplifies assembly in cases or chassis designed for the standard form factor.
An optional heatsink for the Hailo-10H is included in the package . Although the NPU is limited to around 3W, it is generally recommended to install this heatsink, especially if you plan to run intensive workloads for extended periods or demanding benchmarks, as the chip can reach high temperatures.
At the time of the launch, some specialist retailers reported limited stock , a common occurrence whenever a new popular Raspberry Pi accessory is released. Therefore, those wishing to secure a unit in the near future will need to keep an eye on availability from authorized European distributors.
The Raspberry Pi AI HAT+ 2 positions itself as an intermediate solution between the cloud and large AI servers, designed for those who need local processing , data privacy, and controlled costs. It offers a relatively affordable way to combine computer vision and lightweight language models at the edge, with room to grow as the software matures and the catalog of available models expands, provided that the power and memory limitations inherent in a low-power device are accepted.