Any Picture of Text
A phone photo of a page, a screenshot, a scan saved as an image — JPG, PNG and WebP up to 25 MB.
◆ IMAGE TO TEXT
A photo of a page, a screenshot of a form, a receipt on your desk — the words are in there, but you cannot select them. Optical character recognition reads the picture and hands the text back so you can copy, search and edit it.
After reading
217 characters · confidence 95%
Editable, copyable, downloadable as TXT
What the tool does
A phone photo of a page, a screenshot, a scan saved as an image — JPG, PNG and WebP up to 25 MB.
Both language packs are available, and you can tick both when a page mixes them. Turkish letters such as ı, ğ and ş are recognised.
The result lands in a text box you can fix before copying it or saving it as a TXT file, together with a confidence score.
Use cases
Pull numbers and reference codes off a paper document instead of copying them by hand and introducing a typo.
Photograph a page and get a quotable text version for your notes, without a scanner.
Text inside a screenshot cannot be selected. OCR turns an error message or a chat capture into something you can paste and search.
Printed forms and old records become searchable text you can archive properly.
How to get a good result
Recognition quality depends almost entirely on the picture, not on the software. Text that fills the frame, sits roughly horizontal and is evenly lit reads very well — often above 95% confidence. Blur, low resolution, a hard shadow across the page, the curve of a book spine or a photo taken at an angle all cost accuracy.
If a result looks poor, retaking the photo helps more than any setting. Move closer so the text fills the frame, hold the phone parallel to the page, and avoid casting your own shadow on it. The tool shows a confidence score after every read and warns you when it drops below 70%, so you know when to check the output carefully.
Privacy and responsible use
⚠ Good to know
The engine is trained on printed type. Cursive handwriting is generally not readable; neat block capitals sometimes are. Highly stylised display fonts, text printed over photographs and very low-contrast scans are also hard.
Picking more languages than you need slightly slows the read and can lower accuracy, because the engine has to choose between more possible letters. For a purely Turkish document, tick only Turkish; the difference is visible on letters such as ı and ş.
Frequently asked questions
It is free with no account and no daily limit. The only constraint is a 25 MB file size, which is far above a normal phone photo.
The engine and language data are downloaded the first time you use the tool and then cached by the browser, so repeat use is fast. The reading itself never needs a connection because it happens on your device.
Use the PDF OCR tool instead. It renders each page, reads it, and can also build a searchable copy of the PDF with an invisible text layer.
Multi-column layouts and tables confuse line ordering, because the engine reads visually rather than semantically. For a two-column page, cropping one column at a time gives a much cleaner result.
Turkish and English today. Both can be used at once for mixed documents.
Start now
No account, no watermark and no upload. Choose a picture and take the text.
Bubixo document tools are free to use and run entirely in your browser.
This page belongs to PDF and Document Tools, where every tool in the category is listed. These are the ones people usually open next.