Show HN: In-place OCR + Translate for Internet Archive books
Bookxlate is a Chrome extension that uses the Google Gemini API to perform in-place OCR and translation of Internet Archive books, eliminating the need to download images or use separate software. Users must provide their own Gemini API key. The source code is available on GitHub.
Background
- **Internet Archive**: A non-profit digital library with millions of scanned books, many old or out-of-print. Its built-in OCR (extracting text from page images) is often inaccurate and offers no translation.
- **The problem**: Reading a foreign-language book from the Archive normally requires downloading page images, running separate OCR software, then copying text into a translator — tedious and multi-step.
- **Bookxlate**: A Chrome extension that automates this. It captures a page image from the Archive's viewer, sends it to Google's Gemini AI for OCR + translation, and displays the translation directly on the page — no downloads or window-swapping.
- **Requirements**: You need your own Google Gemini API key (a paid AI service). The extension was built mostly with AI-generated code, so reliability may vary.