Photo OCR

Premium

Apple Live-Text style OCR for any image.

About Photo OCR

Photo OCR pulls text out of any image — JPEG, PNG, HEIC, WebP, BMP, TIFF, GIF. Powered by Tesseract's LSTM models with the tessdata_best language packs for maximum accuracy.

Output includes the full recognized text, mean confidence, per-word bounding boxes, and per-word confidence scores. The UI overlays the boxes on the source image so you can click to copy specific words.

How it works

1
Drop or upload an image.
2
Photo OCR processes it locally — no upload to external services.
3
Copy any text, or download the recognized content.

When to use it

Extract text from a receipt photo for expense entry.
Pull a business card into your contacts.
Read text out of a screenshot you can't copy from.
Capture text from a sign or whiteboard photo.

Privacy

Files are processed by the Evixpdf engine in-house with the AGPL-free MIT stack — no third-party cloud upload. Sessions auto-purge after processing.

Frequently asked questions

Short answers to the questions people most often ask about Photo OCR. Read the one that matches your situation — they're written to be skimmed.

1Which languages are supported?

English by default. Drop additional traineddata files into the tessdata/ folder, or use the one-click installer for any pack in the official tessdata_fast / tessdata_best repos.

2Are images uploaded to a third-party API?

No. All OCR runs on the Evixpdf server using local Tesseract models — no third-party cloud calls.

Still stuck?

Browse our hand-written guides or ask us directly — we usually reply within a business day.

Read guides Contact support

OCR Recognition

Add a real text layer to scanned PDFs.

Scan to PDF

Turn a scanned PDF into a fully searchable PDF.

Extract Text

Pull all selectable text out of a PDF as plain text.

Ready when you are

Try Photo OCR now

No signup, no email required. Drag your file in and you're done in seconds.