Some PDFs are hard to work with: scans you can't search, 80-page reports, documents in another language, or a contract that came back with changes. These four tools help you read them, and they all run on your device.
OCR: make a scanned PDF searchable
A scan is a picture of text, so you can't search it or copy from it. OCR PDF reads the text on each page and adds it as an invisible layer in the same place as the words you see, so the pages look exactly the same. Clean, straight scans at 300 dpi read best; handwriting and photos taken at an angle may have mistakes. Recognition runs in your browser, and the language data is downloaded once and kept by your browser.
After OCR, a scan also works with PDF to Excel, PDF to Markdown, Compare PDF and the AI tools below.
Summarise a long PDF with AI
AI Summarizer gives you the key points or a short paragraph. In recent versions of Google Chrome on computers, the summary is written by the browser's built-in AI, on your own device, so your document never leaves your computer; the first time, Chrome downloads the model, which can take a few minutes. In other browsers, GrovePDF picks the document's most important sentences instead.
Translate a PDF and keep its layout
Translate PDF puts the translated text where the original was, with pictures, tables, lines and colours unchanged. Longer translations are made a little smaller to fit their space. Translation uses the translator built into recent versions of Chrome on computers, on your device; each language is downloaded once. Text inside pictures and scans isn't translated, so run scans through OCR PDF first.
Compare two versions of a PDF
Compare PDF checks two versions word by word across all pages, even when text moves to another page. You get a report with both versions side by side: removed words in red, added words in green. The first file is the original; drag the files to swap them. Changes to pictures, colours or layout alone aren't detected.