Tesseract OCR

Health v2 · maintenance only

Adoption not part of the score

76200 stars · 10771 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Tesseract is an open-source OCR engine consisting of the libtesseract library and a command-line program, using an LSTM-based neural network (with legacy engine support) to recognize text in images. It supports over 100 languages, multiple input image formats, and outputs plain text, hOCR, PDF, TSV, ALTO, and PAGE formats.

Use cases

• extract text from scanned pdf documents
• convert images of text to plain text
• ocr receipts and invoices
• recognize text in multiple languages
• add ocr to a document processing pipeline
• generate searchable pdfs from scans

When to choose

• you need a free, self-hosted OCR engine with broad language support
• you want a library or CLI that integrates into document pipelines
• you need structured output like hOCR, TSV, or searchable PDFs

When to avoid

• you need high accuracy on low-quality or complex-layout images without preprocessing
• you need handwriting recognition at state-of-the-art accuracy
• you want a ready-made GUI application

Facets

library · maturity stable

ocr image-processing machine-learning cli computer-vision image-processing pdf developer-tools windows cpp cli cross-platform lstm traineddata hocr multilingual text-recognition natural-language-processing linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="tesseract-ocr/tesseract")