5 million parameters level the scale of billion-level large models: Baidu PaddleOCR surpasses Tesseract to top the GitHub OCR charts

CoinNetwork · 2026-03-30T09:47:56+00:00

Baidu PaddlePaddle's open-source OCR toolkit PaddleOCR has received 73,300 stars on GitHub, surpassing Google's Tesseract to become the highest-starred OCR project. PaddleOCR supports over 100 languages, and recent updates have improved accuracy and performance.

CoinNetwork

2026-03-30 09:47:56

Abstract generation in progress

CoinJie.com news: According to monitoring by 1M AI News, Baidu’s open-source OCR tool library PaddleOCR has surpassed the Google-maintained legacy OCR engine Tesseract (73,200) with 73,300 GitHub stars, becoming the highest-star OCR project on GitHub. MinerU, ranked third, has 57,500 stars. PaddleOCR was open-sourced in 2020, supports over 100 languages, and covers more than 160 countries and regions. Recently, PaddleOCR has undergone intensive updates: the PP-OCRv5 released last week has only 5 million parameters but achieves accuracy comparable to billion-parameter-scale vision-language models on standard OCR benchmarks; PaddleOCR-VL-1.5 set a new record on the document parsing benchmark OmniDocBench v1.5 with an accuracy of 94.5%.

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.