Hailo-10H M.2 AI Acceleration Module

Immersive generative AI experience on the edge

Hailo-10H M.2 AI Acceleration Module

The Hailo-10H is the industry’s first edge AI accelerator to bring immersive generative AI capabilities directly to edge devices. Featuring 40 TOPS of INT4 performance and exceptional power efficiency, the Hailo-10H builds on the success of the market-leading Hailo-8 with a second-generation neural core architecture that is even more powerful and scalable. Hailo-10H includes a direct DDR interface, allowing it to scale for large models such as LLMs, VLMs, Stable Diffusion, and more.

With M.2 form factor, the Hailo-10H accelerator module can be plugged into existing edge devices with an M.2 socket to execute deep neural network inferencing in real-time utilizing low power for a broad range of applications and market segments.

Key Strengths　|　Module Details　|　Tech Specs |　Community Support

Key Strengths

High-performance processing of vision and generative AI models

First to market AI accelerator with generative AI capabilities

Robust and mature software suite, supported by the world’s largest edge AI community

Featuring the Hailo-10H AI accelerator with 40|20 TOPS (INT4|8)

Second generation of Hailo’s market leading AI accelerator

Best-in-class power efficiency; Consumes 2.5W (typical)

Software Architecture

Comprehensive Software Suite

AI software suite which seamlessly integrates with existing deep learning development frameworks to allow smooth and easy integration in existing development ecosystems. The Hailo AI software suite includes:

Explore Hailo models in the Model Zoo and choose the best neural network models for your AI applications

Tech Specifications

Chipset	Hailo-10H
AI Performance	40 TOPS @ INT4 \| 20 TOPS @ INT8
Interface	PCIe3.0 x 4 lanes, M.2 2280 Key-M
Memory	LPDDR4 2GB/ 8GB
Memory Speed	4266 MT/s
Power Consumption	2.5 Watts (Typical)
Supported OS	Windows, Linux, Android
Supported Host Architectures	x86, ARM
Supported AI Framework	Keras, TensorFlow, TensorFlowLite, PyTorch, ONNX
Supported AI Models	Vision Models (> 150+), GenAI Models (LLM, VLM, Whisper)
Package Contents	1 x Quick Start Guide
Operating Temperature	0°C ~ +40°C
Operating Humidity, RH	20% ~ 85%, non -condensing
Storage & Transportation Humidity, RH	10% ~ 95%, non -condensing
Continuous Operating Capability	> 1 week (under normal operating conditions)
Dimension	22 mm x 80 mm
Weight	6 g
Regulation	■ Technical Standard(Global RoHs,China RoHS,EU REACH,J-MOSS) ■ WEEE
Certications	■ FCC ■ CE ■ RCM ■ BSMI ■ VCCI ■ UKCA
Resources	https://community.hailo.ai https://hailo.ai/developer-zone

Benchmarks

Models	Load Time In Sec	TPS	Time To First Token In Sec	Text Time To First Token In Sec	Image Time To First Token In Sec
Whisper-Small	11.92	8.71
Whisper-Base	3.89	23.36
Qwen2-1.5BInstruct-FunctionCalling-v1	7.91	6.23	0.4
Llama3.2-1BInstruct	3.839	8.48	0.49
Qwen2-1.5BInstruct	3.79	8.08	0.32
Qwen2.5-1.5BInstruct	5.05	6.82	0.37
Qwen2.5-Coder1.5B-Instruct	4.76	8.07	0.32
DeepSeek-R1Distill-Qwen-1.5B	4.79	6.98	0.74
Qwen2-VL-2BInstruct	6.226	6.73	0.97	0.32	0.93

All information sourced from Hailo website (retrieved 02/11/26)