Show HN: Textsnap – CPU OCR in <1 seconds/page

Category: cli-tool

Tags: ocr, cpu-inference, paddleocr, onnx, cli, vision-language-model

Score: 6.5/10 (Innovation: 6, Technical: 6, Documentation: 8, Utility: 6)

textsnap is a single-file Python CLI that performs CPU-based OCR on images, screenshots, URLs, and webpages using a two-engine approach: fast PP-OCRv6 for typical cases and a PaddleOCR-VL vision-language model fallback for low-quality or structured content. It features clipboard integration, batch processing, offline operation after model download, and portable model placement for air-gapped use.

Target audience: developers, devops, data engineers, and general users needing local OCR

Repository: https://github.com/kouhxp/textsnap · Python · MIT · 195 stars

View on Hacker News