top of page

Text Recognition & Doc Parsing Toolkit

If you need a multi-page test, please use the AI Studio to support users to upload test files (.pdf file format or image format). The file size should not exceed 20MB, and the number of pages should be controlled within 10 pages. The excess will be truncated. Multi-page document reasoning may take a long time (within 30 seconds), so please be patient.

PaddleOCR 3.0 Architecture

微信图片_20250520090652.png

PP-OCRv5:Universal-Scene Text Recognition Model 

Blank_PPOCRv5.gif

PP-StructureV3:General Document-Parsing Solution

white_v3.gif
bottom of page