Every week, knowledge workers retype text that is already on their screen: figures from a slide deck, a table in a report PDF, a clause in a scanned contract. Capture2Text — free, open-source (GPLv3), portable, Windows-only, by Christopher Brochtrup — collapses that chore into a three-step workflow: capture → proofread → export. This guide sets up the workflow, tunes it for office documents, and shows where it ends and full document tools begin.
Install from the download page (v4.6.3, 2022-03-19, no installer — runs from a folder). New users should skim installation and setup first.
The 3-step workflow
1. Capture with Win+Q
Press the default hotkey Win+Q, drag a box over the text, release. The recognized text lands on your clipboard. That is the entire capture step — two seconds once it becomes muscle memory. For repeating regions (the same report template every week), a fixed-region hotkey is even faster; configure hotkeys in hotkey setup.
2. Proofread (never skip this)
OCR is not perfect, and office documents punish silent errors — a misread digit in a contract figure or a flipped character in a name. Build a 10-second proofread habit: paste the capture into a scratch document, scan for anomalies, fix them against the source image. Common OCR traps: 0/O, 1/l/I, 5/S, and decimal points in small fonts.
3. Export to where the text lives
Paste into Word, Excel, email, a ticket, a CMS — wherever the text needs to go. For longer jobs, capture section by section rather than whole pages; smaller regions OCR more accurately and are easier to proofread.
Once the motion is automatic, the whole loop — see text, Win+Q, glance, paste — takes under ten seconds, which is why heavy users stop thinking of it as OCR at all and start treating the screen as selectable text. The biggest speed gain comes from the proofread step getting shorter with practice: you learn your own most common misreads (the fonts you capture most, the digits your documents use) and start spotting them at a glance instead of reading every word twice.
Office scenarios
- Slide decks (PPT): grab bullet points and chart labels from a colleague’s deck to quote in your own doc — far faster than retyping.
- Reports and dashboards: capture KPI figures from screenshots and BI dashboards that offer no export button.
- Scanned contracts and PDFs: pull clauses out of image-only PDFs. (If the PDF has a real text layer, just select the text — OCR is for when you cannot.)
- Tables into spreadsheets: capture row by row; pasting column-by-column into Excel beats retyping, though complex tables still need manual alignment.
- Error messages and logs: capture an un-copyable error dialog to paste into a support ticket or a search engine.
Accuracy tips for documents
- Zoom before capturing. 150–200% zoom in your PDF viewer or browser gives Tesseract larger, cleaner characters.
- Maximize contrast. Dark text on a light background OCRs best; invert or brighten dark-mode screenshots first.
- Match the OCR language to the document — multilingual docs may need multiple language data files (see adding languages and managing languages).
- Capture at native resolution; avoid photographing screens with a phone when a screenshot is possible.
- Keep the tool updated and portable on a USB stick — portable mode means the same workflow on any work PC with no installation.
From captured text to finished documents
Capture2Text gives you raw text — fast and free. When the job is producing a polished, properly formatted document from that text, dedicated PDF tools take over:
- Wondershare PDFelement (official site) — edit, convert, and assemble PDFs from captured text; the natural “from grab to document” upgrade for heavy office use.
- Adobe Acrobat (official site) — the industry-standard PDF editor for professional document workflows.
The honest division: Capture2Text for quick extraction, PDF editors for authoring. We are an independent site with no affiliation to either vendor; links above may earn us a commission at no cost to you.
Where to go next
- Hotkey setup · Portable mode · Hotkey customization
- Translate feature (for foreign-language documents)
- How to copy text from error dialogs in Windows
- All guides · Download Capture2Text
Frequently asked questions
How do I convert a screenshot to text with Capture2Text?
Press Win+Q, drag a box over the text, and release — the recognized text goes to your clipboard. Then proofread it and paste it where you need it.
Is Capture2Text good for office documents?
Yes for extraction: slides, reports, scanned contracts, and error dialogs. It outputs raw text; for producing polished formatted documents, pair it with a PDF editor like PDFelement or Acrobat.
How accurate is screenshot OCR on small text?
Zoom to 150–200% before capturing, maximize contrast, and match the OCR language to the document. Always proofread — digits and similar-looking characters (0/O, 1/l) are the classic errors.
Can it capture tables into Excel?
Row by row, yes — paste into Excel and align manually. Complex multi-column tables still need cleanup; OCR gives you the characters, not the layout.
Does Capture2Text work offline for this workflow?
Yes. Capture, OCR, and clipboard all work fully offline. Only the optional translation feature needs internet.
Is Capture2Text really free for commercial use?
Yes — it is free and open source under GPLv3 by Christopher Brochtrup, with no paid tier and no feature restrictions.