Believe it or not, some people still print documents to physical pieces of paper. Optical Character Recognition (OCR) software takes those printed documents and converts them right back into machine-readable text. We’ve found some of the best free OCR tools and compared them for you here.
No OCR program is perfect, so you’ll have to double check the results and fix a few problems. Still, it’s a lot faster than typing the entire document back into the computer. Each of these free OCR software tools has its own strengths, and all of them will get the job done.
The Methodology
To compare these tools, I printed out MakeUseOf’s About page and scanned it back into the computer. With multiple columns and imperfections, this is not an ideal document to OCR. But that’s a good thing – real-life OCR operations will often be with imperfect documents.

Google Docs
Google Docs has integrated OCR support. It uses the same OCR engine that Google uses to scan books and understand text in PDF files.
To get started, open the Google Docs website and start uploading a file. You can’t scan directly from your scanner into Google Docs; you’ll have to scan the document as an image or PDF file first. If you don’t have a scanner, you can try scanning a document with your smartphone’s camera.

Enable the “Convert text from PDF and image files to Google documents” check box when you upload the file.

After you upload the file, it will appear as a new text document in Google Docs.
Google Docs did a pretty good job here. It struggled to understand the web addresses, but all these tools did.

Unlike many free online OCR tools, like Free OCR (which is different from the FreeOCR below), Google Docs doesn’t have any limits on the amount of pages you can upload.
FreeOCR
FreeOCR is a simple, easy-to-use frontend for the open-source Tesseract OCR engine, originally developed by HP Labs. FreeOCR comes bundled with Tesseract, so you don’t have to install anything extra. We’ve mentioned FreeOCR in the past.

Use the Scan button to scan a page directly from your scanner or use the Open option to open an image or PDF file.
Click the red X button to clear the pre-filled text before continuing. After opening the file, click the OCR button on the toolbar and FreeOCR will start crunching away.

FreeOCR gave good results on the main block of text, but it was confused by the columns. To fix this, I selected the block of text I wanted and used the “Crop image to selected area” option.

The results were much better after running the OCR operation on a specific section of the document.

Cuneiform OpenOCR
Cognitive Technologies developed Cuneiform as a commercial OCR solution, but eventually released it as freeware. Cuneiform OpenOCR has an unpolished interface, but there’s an excellent OCR engine underneath.
The download page is in Russian — scroll down and click the “english version” setup link to download and install Cuneiform.

After you install it, you’ll find that it didn’t create appropriate Start menu entries. The “NewShortcut6” shortcut will launch OpenOCR.

From the File menu, use the Open option to open a file or the Scan option to scan a document.

After it’s in OpenOCR, use the Recognize option in the Recognition menu to OCR the text. Cuneiform OpenOCR will save it as a file.

Cuneiform OpenOCR provided good text recognition. It also preserved the formatting and text size differences – unlike the other programs here.
The Verdict
Each of these programs has its strengths.
- Google Docs can OCR documents without downloading anything to your computer.
- FreeOCR offers a very easy-to-use interface, but requires some fiddling if a document contains columns.
- Cuneiform OpenOCR preserves much of the original document’s formatting, but its interface lacks polish.
Once you’re done with the OCR process, you may want to spell-check your document. Depending on your use, you may not even have to OCR documents at all – you can convert a paper book to an ebook without OCRing it.
Which OCR software works best for you? Do you have a different favorite OCR program that we didn’t mention here? Leave a comment and let us know.
Image Credit: Woman’s Hand With Working Copier via Shutterstock
MakeUseOf Recommends
More articles about:
Hide 32 Comments
Does it support chinese ?
I believe that Google Docs does. Tessaract (FreeOCR) also seems to support Chinese. Cuneiform may — I’m not sure.
But one of these solutions should definitely work for you!
Thank you very much, it is useful.
Google Docs while not having a page limit does have:
Sorry, this file is too big. We can only convert files up to 2 MB in size.
:(
Thanks for sharing that, Danny!
I’m sure all services have a max file size — it’s sad that Google Docs’ file size is so low.
However, you could always split the document into several files and upload them one by one — Many free services say “x pages” or “x documents” as a limit; that’s no fun.
I use ABBYY Finereader Sprint free version – works fine for home/personal use.
Thanks, Gordon — why did you choose it? What makes it better than the other options to you?
Hi! I also use Sprint but, probably, for big projects with a lot of formatting and complex layout I need a full product that will save the original structure and convert pdf. Not sure what product to buy, do you know if I can upgrade from Sprint (and if it is cheaper) to complete version of finereader ?
I’m not sure what the best paid product to buy is either — you might try asking on MakeUseOf Answers to get more feedback: http://www.makeuseof.com/answers/
I’m sure you can upgrade to the full version though — since it’s just a matter of paying for the full version and downloading it.
The best two packages bar none for accuracy, layout preservation etc (not necessarily for speed or ease of use) are Abbyy FineReader Pro and Nuance OmniPage Professional. Personally, I prefer the Abbyy product which is much more elegant and flexible (allowing you, for example, to define specific character sets which can enhance the accuracy of the document). However, the Nuance product is also very good.
Another big player is IRIS, whose products are are geared towards “productivity” rather than for accuracy. They are much faster, and most of the IRIS packages are largely “hands off” (no interactive layout recognition or spell checking, for example). This is suitable for those who are scanning into searchable PDFs, where absolute accuracy is not paramount, but speed is!
What? No OneNote? It OCRs all pasted pictures by default, so you can easily find them. I don’t use Evernote but expect it to be similar.
I’ve heard good things about OneNote and we’ve covered it in the past, but I wanted to cover free OCR tools here. Many people may already own OneNote, but it is a commercial product that costs money to use.
I use ocr a lot and have tried almost every free option out there. For straight forward text they handle just fine, but send through complex layouts and expect to spend hours editing. Recently I forked out for the premium Omnipage, and although I am now broke, I have to admit that the program makes everything else look inefficient. Simple pages are ocr-ed in seconds and are perfect. Complex layouts take a little longer because you need to define graphic, text and table areas, but the results are near perfect. I am not saying that the options mentioned here can’t do the job, but if you have big projects in need of ocr, or ocr documents regularly, then save the headache and pay for the master!
Thanks for sharing your experience, Carie. I’ve never used Omnipage so I can’t speak to it, but it may well be a good idea to pay for more powerful OCR software if you need it.
Does OmniPage read handwritten text?
Apparently it isn’t designed for handwritten text, so it may not be the ideal solution. Here’s the official answer to your question:
http://nuance.custhelp.com/app/answers/detail/a_id/828/~/does-omnipage-recognize-handwritten-text%3F
Does OminPage read handwritten text?
The OCR tools seems to be great.
Thanks, Mark! I tried to choose a variety of tools with a variety of engines and strengths.
Hi Chris Hoffman,
I like Cuneiform OpenOCR,
but Cuneiform OpenOCR is not support to do OCR on PDF files
do you know how to make coneiform openocr to do ocr on pdf files?
thanks
Looks like you’ll have to convert the PDF to an image file, then you can open the resulting image file in OpenOCR.
Here are some ways to do it: http://www.makeuseof.com/tag/extract-images-pdf-files-save-windows/
Not sure whether Google Docs ever had this same limitation, but Google Drive (which it appears Google plans to eventually replace Google Docs with) limits its OCR’ing of any PDF you upload to just the first ten pages. The 2 MB file size limit also still applies, but this ten-page limit is even more restrictive.
However it’s good they still don’t restrict the number of documents you upload. As a result, you can still OCR an arbitrarily large number of pages for free on Google Drive — ten pages at a time.
That’s unfortunate. It’s still a pretty great service, though — hard to complain when it’s free and happens on their servers.
I agree Chris. Also note, since the ten-page restriction applies only to PDF format there could be a simple workaround. There might be another multipage file format such as TIFF that a person could convert their PDF to before uploading.
Thanks, great tip!
It’s clear that Google Docs is supposed to be a very basic OCR tool, not an industrial-strength one.
I need a program to scan & read handwritten and digital text, convert it to pdf, searchable and networkable.
Handwritten text is a much harder problem, as handwritten characters aren’t as distinct as typed ones. It’s difficult to find a program that will work well for this.
You might try SoftWriting, which has a free trial: http://www.charactell.com/SoftWriting.html
You might also try SimpleOCR, which seems to have this feature — but it doesn’t seem to work very well: http://www.simpleocr.com/Download.asp
I like to professionally use OCR (50 pages/day), but i don’t like to spend thousands of dollars; what is a good program to buy? Is Omnipage the best option?
Thanks, Frank
I haven’t used any commercial OCR programs myself, so I don’t feel comfortable recommending any specific ones.
You should ask this question in MakeUseOf Answers: http://www.makeuseof.com/answers/
The question will get more attention and people with experience can give you a better answer!
GT Text is also very good.
It supports Chinese (just downloading the language in preferences)
Japanese and a myriad of languages
http://gttext.googlecode.com
Thanks for sharing. It’s open source, which is pretty awesome.
Does any of these Open Source software’s have scheduling feature. Because OCR’ing big files and numerous of them take long time. So I am thinking of scheduling them. The feature that ABBY Prof.Edition has?
Thanks,