In 2023, the trend of large models surged, with many large models from both domestic and international sources opening for internal testing. This article provides a comprehensive horizontal evaluation of four major large models: ChatGPT, Wenxin Yiyan, Tongyi Qianwen, and Xunfei Spark. Tests and assessments were conducted in areas such as multi-modal capabilities, voice interaction abilities, text-to-image/video/audio capabilities, and knowledge-based question-answering abilities.