Pyyan / OCR & Document AI / Tesseract

Tesseract

The 1985 engine that ran OCR for a generation.

currentGoogle ↗Verified 2026-08-13

1985

first released

Specifications

OmniDocBench	~55as of 2026-08-15 · low confidence
Open weights	Yes
Handles	Clean printed text
Licence	Apache 2.0
Kind	Classical engine

Official links

Google official site

What matters

Included here as the baseline everything is measured against
Struggles badly with layout, tables and anything skewed

Best for clean scans, and nothing else.

Compare with

All 5 side by side →

Tesseract vs GLM-OCR →Tesseract vs dots.ocr →Tesseract vs Qwen3-VL →Tesseract vs DeepSeek-OCR →Tesseract vs Mistral OCR →

Also in OCR & Document AI

GLM-OCR · Zhipu AI dots.ocr · Xiaohongshu Qwen3-VL · Alibaba DeepSeek-OCR · DeepSeek Mistral OCR · Mistral AI