NexaAI/yolo26x-npu

YOLO26x

Model Description

YOLO26x is the extra-large variant of the YOLO26 model family, built to deliver maximum detection accuracy for demanding and large-scale computer vision workloads.

As the most powerful model in the YOLO26 lineup, YOLO26x offers significantly increased capacity, making it well-suited for complex scenes, small-object detection, and high-resolution inputs. It retains the end-to-end, NMS-free architecture, ensuring a consistent and streamlined inference workflow across deployment environments.

YOLO26x is pretrained on the COCO dataset and is intended for scenarios where accuracy is the primary objective and compute resources are readily available.

Quickstart

1. Install NexaSDK and create a free account at sdk.nexa.ai
2. Activate your device with your access token:

   nexa config set license '<access_token>'
   

3. Run the model on Qualcomm NPU in one line:

   nexa infer NexaAI/yolo26x-npu
   

Features

• Extra-large model size for highest detection accuracy
• End-to-end detection with NMS-free inference
• Optimized for accuracy-first deployments
• Pretrained weights available out of the box
• Ultralytics ecosystem support for training, validation, inference, and export

Use Cases

• High-accuracy object detection systems
• Large-scale video analytics
• Autonomous driving and advanced robotics
• Industrial and scientific vision applications
• Research and benchmarking

Inputs and Outputs

Input:

• Images or video streams, automatically preprocessed by the Ultralytics framework

Output:

• Bounding boxes
• Class labels
• Confidence scores

License

This repo is licensed under the Creative Commons Attribution–NonCommercial 4.0 (CC BY-NC 4.0) license, which allows use, sharing, and modification only for non-commercial purposes with proper attribution. All NPU-related models, runtimes, and code in this project are protected under this non-commercial license and cannot be used in any commercial or revenue-generating applications. Commercial licensing or enterprise usage requires a separate agreement. For inquiries, please contact [email protected]