FAQ · STRAIGHT ANSWERS
Frequently asked questions
Real questions people have asked, answered in plain language. If yours isn't here, the how it works page or an email to [email protected] will sort it out.
Is this image to text converter really free?
Yes. No account, no trial, no credit card, no "sign up to unlock the full version" step. I'm not going to promise it stays free forever — running a converter costs real money — but if pricing ever shows up, the homepage will say so plainly rather than sneaking it into a settings page.
What happens to my files after I convert them?
Your file is sent over an encrypted connection, the text is extracted, and the file is deleted from the server the moment the task finishes. Nothing is kept, cached, or reused. Want extra control? Use the mask tool to blur sensitive parts before the file ever leaves your machine.
Which image formats are supported?
More than 30 — JPG, PNG, WEBP, HEIC, TIFF, GIF, BMP, AVIF, JPEG 2000, ICO, PSD, plus several camera RAW formats like CR2, NEF, ARW, DNG, ORF, and RW2, and SVG. The formats page lists every single one.
Can I convert a PDF to text?
Yes. Single and multi-page PDFs both work. A scanned PDF is taken apart page by page, each page goes through OCR, and the text comes back in reading order. Old manuals and archive scans become searchable this way.
How many files can I convert at once?
Up to 10 files per batch, up to 20 MB each. Every file keeps its own status and its own result box, so a folder of nine screenshots and one photo all convert without you queueing them one at a time.
How accurate is the extracted text?
It depends more on the image than on the engine. Clean, flat, well-lit printed text usually converts close to perfectly. Glare, shadows, off-angle photos, and very low resolution cause mistakes. There's a practical guide to what helps in the accuracy article.
Can I convert handwriting?
Sometimes, and I'd rather be honest about it than promise magic. Neat, blocky print can come through. Cursive is mostly a miss. OCR engines are built around printed characters, and handwriting recognition is its own field. Always glance at the output before relying on it.
Why does the text come back in the wrong order?
Usually one of two things. Newspaper-style multi-column layouts confuse the reading order, and a photo shot from an angle can make lines blur together. Straightening the photo or flattening the page usually fixes it. If it's a two-column layout, a crop of one column at a time helps.
Can I hide sensitive content before uploading?
Yes — every file has a Mask button. It opens a small editor where you can paint over, box out, or blur anything sensitive: names, card numbers, addresses, serials. Only the masked version is uploaded, so the covered content is gone before it ever leaves your machine.
Which languages are supported?
Over 100, covering Latin, Cyrillic, Arabic, Hebrew, CJK, Indic, and Southeast Asian scripts. Pick a language from the dropdown or leave it on auto-detect. Manual selection is more reliable for short snippets; the languages page has the honest details.
Why does the result say "no text detected"?
The engine looked at the file and didn't find shapes it could confidently call letters. Common causes: an image that's mostly empty space, a thumbnail-size screenshot, or heavy handwriting. A sharper, closer crop of the text usually fixes it.
What exactly is OCR?
Optical character recognition. Software looks at the shapes of printed characters in an image and converts them into plain, editable text. That's the entire trick behind every image to text converter, this one included.
Can I use the extracted text commercially?
Yes. There are no strings attached to the output. Text extracted from your own documents is your text, and you can use it however you like — research, work, publishing, whatever.
Is there an API?
Not a public one yet. If you want to run OCR from a script or an app, write to [email protected] and describe the use case. Enough people ask and it becomes a feature instead of a wish.