OpenCV 5 has been released as the biggest update in years. It includes a new DNN engine, enhanced ONNX support, hardware acceleration improvements, and better Python integration.
OpenCV 5 has been released. This release brings a new DNN engine, enhanced ONNX support, hardware acceleration improvements, better Python integration, new data types, expanded 3D vision capabilities, improved documentation, and a cleaner architecture. The pip version will be released on June 8, 2026.
OpenCV has been the foundation for computer vision research, robotics, embedded vision, AI applications, industrial inspection, AR/VR, medical imaging, and countless production systems for over two decades. It now has over 86,000 GitHub stars and more than a million installs per day. Since OpenCV 4, computer vision has changed dramatically with deep learning, transformers, large vision models, edge deployment, heterogeneous hardware, and Python-first workflows. OpenCV 5 was designed to meet that reality.
OpenCV 5 is a major leap forward, not just an incremental update. Enhanced deep learning model support will make it easier to deploy state-of-the-art models, and hardware acceleration improvements are expected to boost performance on edge devices. Better Python integration will significantly improve the developer experience. This will have a substantial impact across the entire computer vision ecosystem.
Regarding the OpenCV 5 release, the HN community discussed the appropriate division of roles between traditional computer vision methods and modern AI models. Some argued that OpenCV should wrap VLMs or image generation models beyond existing approaches like YOLO, but most countered that traditional CV methods remain essential given the limited resources of edge devices and real-time processing requirements (under 300ms), and that switching to AI models is not realistic in terms of performance and efficiency. A practical perspective stood out, especially for specific simple tasks like industrial pick-and-place or orange sorting, where AI models introduce unnecessary complexity.