1: The Baidu PaddlePaddle team has launched PP-ChatOCR, a general image key information extraction tool based on the Wenxin large model. 2: PP-ChatOCR combines OCR text recognition and large model technology to extract key information from images in various scenarios. 3: PP-ChatOCR has been launched on PaddleX, supporting high-performance deployment, and has achieved good accuracy and stability.