Search the text inside your images

Why normal search can't find it

To your operating system, a screenshot is just pixels named something like Screenshot 2026-03-14.png. The words rendered inside it — the phone number, the order total, the heading on a slide — are invisible to file search. That's why you end up scrolling through hundreds of thumbnails looking for "the one with the address on it."

What OCR changes

OCR (optical character recognition) reads the text rendered in an image and turns it into actual, searchable words. Once PicSy has extracted that text, it's indexed right alongside each image's AI description and tags — so a plain search matches it.

Things you can suddenly find

• The screenshot with a specific phone number or error code.
• A receipt or invoice by the vendor name or amount.
• A slide or whiteboard photo by a heading or keyword.
• A meme by its caption, or a book page by a quote.

How to use it

• 1. Add the folders with your screenshots and photos so they're indexed.
• 2. Extract text. Run analysis (or the dedicated "extract text" action) so the words inside each image are captured.
• 3. Just search. Type any word or phrase you remember seeing — results include images that contain it.

Private by default

Text extraction uses whichever AI model you choose. Pick a local model (via Ollama or LM Studio) and it runs entirely offline — ideal for receipts and anything sensitive. Prefer a cloud model? Only the image you analyze is sent, using your own key. Learn more on the features page or download PicSy.

FAQ