V2X-Real is a large-scale dataset for Vehicle-to-Everything (V2X) cooperative perception, containing 33K LiDAR frames, 171K camera images, and over 1.2M annotated 3D bounding boxes of 10 categories. It was collected using two connected automated vehicles and two smart infrastructures equipped with multi-modal sensors, including LiDAR and multi-view cameras. The dataset includes four sub-datasets for Vehicle-Centric, Infrastructure-Centric, Vehicle-to-Vehicle (V2V), and Infrastructure-to-Infrastructure (I2I) cooperative perception. It is designed to support multi-class multi-agent benchmarks for V2X cooperative perception, providing a comprehensive framework for research in this area. The dataset is collected in urban and freeway environments, featuring a high density of traffic and vulnerable road users, offering challenging scenarios for V2X cooperative perception. The dataset includes detailed sensor specifications, data acquisition, annotation, and processing pipelines, as well as evaluation metrics and benchmark results for various cooperative perception methods. The V2X-Real dataset and benchmark codes will be released to facilitate further research in V2X cooperative perception.V2X-Real is a large-scale dataset for Vehicle-to-Everything (V2X) cooperative perception, containing 33K LiDAR frames, 171K camera images, and over 1.2M annotated 3D bounding boxes of 10 categories. It was collected using two connected automated vehicles and two smart infrastructures equipped with multi-modal sensors, including LiDAR and multi-view cameras. The dataset includes four sub-datasets for Vehicle-Centric, Infrastructure-Centric, Vehicle-to-Vehicle (V2V), and Infrastructure-to-Infrastructure (I2I) cooperative perception. It is designed to support multi-class multi-agent benchmarks for V2X cooperative perception, providing a comprehensive framework for research in this area. The dataset is collected in urban and freeway environments, featuring a high density of traffic and vulnerable road users, offering challenging scenarios for V2X cooperative perception. The dataset includes detailed sensor specifications, data acquisition, annotation, and processing pipelines, as well as evaluation metrics and benchmark results for various cooperative perception methods. The V2X-Real dataset and benchmark codes will be released to facilitate further research in V2X cooperative perception.