Free OCR tools
Turn photos, screenshots and scanned pages into text you can copy, edit and search.
OCR, optical character recognition, reads the letters in a picture and gives them back as real text. Choose the tool by what you want at the end: text to copy, a table to calculate with, or a scanned PDF that you can search.
Many languages are supported, including Urdu, Arabic and Chinese. Setting the language yourself gives the best accuracy.
OCR tools · 3 tools
Choosing the right OCR tool
OCR — optical character recognition — turns pictures of text into real text. Image to Text reads photos, screenshots and handwriting and gives you text to copy, a Word file or a searchable PDF. Image to Excel rebuilds a table from a photo or screenshot into a spreadsheet. PDF OCR adds a hidden text layer to a scanned PDF so it can be searched and copied while looking exactly the same.
Urdu, Arabic, Chinese and many other languages are supported; setting the language explicitly gives the best accuracy.
For the cleanest result, photograph the page flat and straight on, in even light, with the text filling the frame. Screenshots are read almost perfectly because the text is already sharp. Handwriting works best when it is neat; any word the reader was unsure of can be flagged so you know exactly what to check.
Which one to use: text you want to copy or edit — Image to Text; a table you want to calculate with — Image to Excel; a scanned document you want to keep as a PDF but search — PDF OCR. For a digital PDF (not a scan), PDF to Text and PDF to Excel under PDF tools are faster still.
Good to know
Always check numbers, names and anything you plan to sign or file. OCR is very accurate on clear print, but a smudge can turn a 5 into an S or an 8 into a B, and in a table a single misread digit changes the total. Image to Excel keeps each value in its own cell, so a quick scan down each column catches most mistakes.
A photo taken from an angle, in shadow or out of focus gives the reader far less to work with. Retaking the picture usually helps more than any setting, and a scanner app that flattens the page before saving helps more still.
Multi-page documents are easiest as a single PDF. Put the scans together with JPG to PDF, then run PDF OCR once, and you get one searchable file instead of a folder of separate photos. If a document mixes languages, such as English and Urdu on one form, leave PDF OCR on Auto-detect, or pick both in Image to Text, so neither is read as the other. The finished file can then be searched with Ctrl+F like any digital document, and copied from without retyping a word.
Guides and how-tos
Answers to the questions that come up most, from the people who build the tools.
Questions about the OCR tools
How do I get the best OCR result?
Photograph the page flat and straight on, in even light, with the text filling the frame. Screenshots are read almost perfectly.
Can it read handwriting?
Neat handwriting works best. Words the reader was unsure of are flagged so you know what to check.
Should I check the output?
Always check numbers, names and anything you plan to sign. A smudge can turn a 5 into an S.
Can I make a scanned PDF searchable?
Yes. PDF OCR adds a hidden text layer, so the file looks the same and can be searched and copied from.
More free tools
Every category works the same way: no signup, and your files are deleted an hour after processing.
33 free tools in 8 categories — no signup needed