Cookie Settings
By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage and assist in our marketing efforts. More info
Cookie Settings
By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage and assist in our marketing efforts. More info
Accept All Cookies
Cookie Settings
Yolo Vision Shenzhen
Shenzhen
Join now
Back to partners

High-performance edge AI optimized for efficiency with Axelera AI

Ultralytics partners with Axelera AI to deliver smart inference at the edge with Axelera's Voyager® SDK running on Metis® AI Processing Units.

Request license
Explore our solutions
High-performance edge AI optimized for efficiency with Axelera AI

About Axelera AI

Axelera AI is delivering the world’s most powerful and advanced solutions for AI at the Edge. Its industry-defining Metis® AI platform – a complete hardware and software solution for AI inference at the edge – makes computer vision applications more accessible, powerful and user friendly than ever before.

Based in the AI Innovation Center of the High Tech Campus in Eindhoven, The Netherlands, Axelera AI has R&D offices in Belgium, Switzerland, Italy and the UK, present in 20 countries with customers on every continent. Its team of experts in AI software and hardware come from top AI firms and Fortune 500 companies.

Why choose Axelera for Ultralytics YOLO?

Deploy Ultralytics YOLO models with unmatched performance and efficiency

Optimized for Ultralytics YOLO

Full YOLO support out of the box, from YOLOv8 to YOLO26 without manual configuration needed.

Edge-native performance

Metis AIPUs and D-IMC technology: Exceptional edge throughput, minimal power consumption.

Real-time inference

On-device inference across all five YOLO tasks, with real-world speed and reliability.

Lower cost of ownership

A fraction of the power cost of GPUs, with none of the performance trade-offs.

Easy integration

Single-step export via the Ultralytics Python package. Voyager SDK does the rest.

Future-proof

Full Ultralytics YOLO family support, with new architectures integrated as they release.

The complete solution

Train with Ultralytics YOLO

Build and refine your custom YOLO models on Ultralytics Platorm. From Ultralytics YOLOv8 to YOLO26, access state-of-the-art architectures, curated datasets, and powerful optimization tools, all designed to get you to production faster.

Export to the Axelera format

Convert and quantize your trained YOLO model for Axelera Metis AIPUs deployment in a single command. The export process compiles your model into an optimized .axm package, including network topology, weights, and metadata.

Deploy on Metis AIUPs

Run inference on Axelera's Metis AI Processing Units with high throughput and ultra-low power consumption. With four independently programmable cores per chip, you can run up to four models in parallel without needing PyTorch at inference time.
Export to the Axelera format

Technical integration

Seamless integration between Ultralytics models and Axelera AI hardware
Supported models
Supported tasks
Performance benchmarks
Usage example

Ultralytics YOLOv8

8

YOLOv8 improved performance and added supported tasks: Segmentation, classification, and pose estimation.

YOLOv8n
YOLOv8s
YOLOv8m

+22

Ultralytics YOLO11

11

YOLO11 supports the computer vision tasks users are familiar with while offering better accuracy, precision, and speed.

YOLO11n
YOLO11s
YOLO11m

+22

Ultralytics YOLO26

26

YOLO26 is a lighter, smaller, and faster model with the nano variant delivering up to 43% faster inference on standard CPUs.

YOLO26n
YOLO26s
YOLO26m

+22

Object detection

Detect objects in an image or video with axis-aligned bounding boxes.

Instance segmentation

Create pixel-precise masks highlighting objects within a frame.

Image classification

Assign one or multiple labels to an entire image.

Pose estimation

Detect and label body keypoints for body position monitoring.

Oriented bounding boxes object detection

Detect rotated objects with angle-aware bounding boxes.
Two warehouse workers wearing yellow safety vests and helmets inspecting stacked wooden pallets.
Purple grapes on a vine with individual grapes outlined, demonstrating instance segmentation.
Bottles on a production line labeled “No defects,” indicating automated defect detection in manufacturing.
Zebra near water with keypoints overlaid on its body, demonstrating pose estimation.
Aerial construction site with trucks and excavators highlighted by bounding boxes, showing object detection and tracking.

Real performance on Metis AIPUs

Learn more about Ultralytics YOLO’s performance on Axelera hardware across FPS, latency, accuracy, tasks and model sizes.

Read documentation
Deploy on Metis AIUPs

Export to Axelera Format

Python
CLI
from ultralytics import YOLO
# Load a YOLO26 model
model = YOLO("[yolo26n.pt](http://yolo26n.pt)")
# Export to Axelera format
model.export(format="axelera")  # creates 'yolo26n_axelera_model' directory

yolo export model=yolo26n.pt format=axelera

Inference with Axelera Model

Python
CLI
from ultralytics import YOLO
# Load the exported Axelera model
model = YOLO("yolo26n_axelera_model")
# Run inference
results = model("https://ultralytics.com/images/bus.jpg")
# Process results
for r in results:
    print(f"Detected {len(r.boxes)} objects")
    [r.show](http://r.show)()  # Display results

yolo predict model='yolo26n_axelera_model' source='https://ultralytics.com/images/bus.jpg'

Get started with Axelera AI

Everything you need to deploy YOLO models on Axelera AI’s Metis processing units
Complete integration guide for deploying Ultralytics YOLO on Axelera AI’s untis

Get started with Axelera AI

Everything you need to deploy YOLO models on Axelera AI’s Metis processing units

Learn more
Learn more about real use cases featuring Ultralytics and Axelera AI’s solutions.

Axelera AI case study

Learn more about real use cases featuring Ultralytics and Axelera AI’s solutions.

Learn more
Explore how to deploy Ultralytics YOLO models directly to the Metis® AIPU.

Axelera AI blog

Explore how to deploy Ultralytics YOLO models directly to the Metis® AIPU.

Learn more

Become an Ultralytics partner

Join our partner ecosystem and unlock new opportunities to deliver cutting-edge AI solutions

Become a partner