OCR Online: Extract Text from Images & Scanned PDFs
Turn a photo, scan, or PDF into searchable, editable text. Scanquil reads text two ways: a free on-device engine for everyday scans, and an optional AI-powered pass for the harder cases -- skewed photos, dense tables, faint receipts, and handwriting -- the free engine struggles with.
Free, on-device OCR
The default engine reads text directly in your browser -- nothing is uploaded, and it keeps working offline once the page has loaded. It's accurate on clean, well-lit scans and typed documents, and it's what every tool below uses unless you specifically pick an "AI" variant.
AI OCR for harder scans
When a photo is skewed, low-contrast, mostly handwriting, or the page is dense with small print, the free engine's accuracy drops. These send just that one image to a cloud OCR provider for a meaningfully better read, for a small pay-per-use fee shown on the button before you tap it -- no subscription, no stored balance.
After OCR: search, edit, and ask
Every scan is saved into a group, where its text is searchable and editable, and it's exactly what Ask AI answers questions from -- with clickable citations back to the page an answer came from. Export any group as a single backup .zip at any time.
Need structured data, not just text?
For pulling tables, form fields, and invoice line items out as structured data instead of plain text, see Document AI.
Privacy
Free OCR never leaves your device. AI OCR sends only the one image or page you choose to process -- see exactly what each tool sends, tool by tool, on the privacy page.
Frequently asked questions
Is OCR really free?
Yes -- the on-device engine behind Image to Text, PDF to Text, Image to Word, and PDF to Word has no cost, no account, and no upload, ever. Only the AI-powered variants for harder scans are pay-per-use.
Does OCR work offline?
The free on-device tools do, once the page itself has loaded -- no file leaves your device. The AI-powered tools need a connection, since they send one image to a cloud provider for that single request.
Can OCR read handwriting?
The free engine is built for printed text and struggles with handwriting. The AI-powered tools, like Image to Text (AI), read handwriting meaningfully better, though accuracy still depends on legibility.
What file types can I run OCR on?
Photos (JPG, PNG, WebP) and PDFs, including multi-page and scanned PDFs -- pick whichever tool above matches your input and the output format you want.