Tesseract

Tesseract is an open-source optical character recognition (OCR) engine, originally developed at HP and now maintained by Google along with a community of contributors. It ships as a command-line program and a library, libtesseract, and supports both a newer LSTM neural-network-based recognition engine (the default since Tesseract 4) and the legacy pattern-matching engine from Tesseract 3 for compatibility.

It reads common image formats like PNG, JPEG, and TIFF, recognizes text in more than 100 languages out of the box using Unicode/UTF-8 support, and can output plain text, searchable PDF, hOCR, TSV, and other structured formats - though as with any OCR engine, result quality depends significantly on input image quality.

Tesseract is a fit for teams that need to extract text from scanned documents, photographed forms, or images as part of a data pipeline - digitizing paperwork, indexing scanned archives, or feeding text into an LLM/RAG pipeline - complementing PDF.js (also in this catalog), which renders and parses PDFs that already contain a text layer, rather than recognizing text from pixels.