Microsoft Lens is gone. Scan to editable Word, free, 145 languages.
Drop a phone photo, screenshot, or scanned PDF — get a real .docx you can edit, search, and quote. Enterprise-grade OCR plus an optional AI proofread pass that fixes the typical digit-letter confusions cheap OCR tools leave behind. No watermark, no signup, no per-page paywall.
Drop an image or scanned PDF here
JPG / PNG / HEIC / WEBP / TIFF / BMP / PDF
Scanned paper into a real Word document, in seconds
Microsoft Lens was retired on January 9, 2026, and the Copilot replacement no longer saves to OneNote, Word, or PowerPoint. Free Web OCR alternatives either throttle to a few European languages, hide useful features behind a paywall, or watermark the output. ExcelMaster Word OCR closes that gap. Drop a photo, screenshot, or scanned PDF; we recognise the text in 145 languages, optionally pass it through a state-of-the-art language model that fixes the typical OCR slips, and assemble a clean .docx you can open in Word, Pages, or Google Docs and edit immediately.
What you get with one click
True .docx, not an image
The output is a real Word document. Every word is selectable, copy-pasteable, searchable, and editable in any modern word processor.
145 languages including CJK and RTL
Recognise Chinese, Japanese, Korean, Arabic, Hindi, Cyrillic, Greek, Hebrew, Thai, Vietnamese, and dozens more, all in one pass.
Optional AI proofread
Toggle on to send the OCR text through an enterprise-grade language model that fixes digit/letter confusions, restores broken date and currency formats, and rejoins fragmented paragraphs.
Mobile camera, one tap
On phones, the camera button opens the rear camera directly. Snap a textbook page, board, receipt, or contract and get a Word file before you put your phone down.
Encrypted PDFs accepted
If your scanned PDF has a password, paste it once and we open + recognise + return a Word file with no password attached. No "unlock first, convert second" detour.
Auto-deleted after 24 hours
Files come in over HTTPS and are wiped from disk within 24 hours. We never train models on your uploads.
How it works
- 1Drop or snap your image / scanned PDF (JPG, PNG, HEIC, WEBP, TIFF, BMP, PDF).
- 2Pick the OCR language (or leave on Auto-detect) and the output format (.docx by default).
- 3Optionally turn on the AI proofread pass for the cleanest output - adds a few seconds per page.
- 4Click Convert to Word and download the .docx. Open it in Microsoft Word, Pages, or Google Docs.
Why ExcelMaster Word OCR beats the free alternatives
145 languages, free, no paywall
iLovePDF locks OCR behind Premium and only ships 14 European languages. We keep all 145 free, including Chinese, Japanese, Korean, Arabic, and Hindi.
AI proofread pass nobody else ships
Most online OCR tools dump raw recognition output. We optionally run an LLM proofread that fixes the predictable OCR errors and restores paragraph flow.
Mobile-first, like Lens used to be
Native rear-camera capture, HEIC accepted, designed for phones. Microsoft Lens users will recognise the workflow on first use.
No 5-files-per-hour throttle
OnlineOCR.net caps free users at 5 files per hour and 15 MB. We don't. Convert as many scans as you need.
Honest privacy
Files travel over HTTPS, run through a hosted OCR + LLM stack, and are deleted within 24 hours. We say so plainly - no fake "never leaves your device" marketing.
Independent benchmark you can audit
Our /benchmarks/word-ocr-2026 page publishes the dataset, methodology, and per-engine numbers so you can verify the claims.
Frequently asked questions
Is this really free?
Yes - free, no signup, no watermark. We monetise our paid Excel agent. No per-page paywall and no 5-files-per-hour throttle.
Is this a proper Microsoft Lens replacement?
For the "scan a page -> editable Word" workflow, yes. Whiteboard perspective correction and business-card-to-contact extraction land in a later release.
Which languages does the OCR cover?
145, drawn from the full Aspose.OCR matrix, including Chinese (simplified + traditional), Japanese, Korean, Arabic, Hebrew, Hindi, Thai, Vietnamese, Cyrillic, Greek, and the full Latin set.
What does the AI proofread pass do?
Four corrections in one round-trip: digit-letter confusions, restore split numbers and dates, locale-aware date and currency formats, rejoin fragmented paragraphs. Strict-JSON output - never invents content.
How big a file can I upload?
Single uploads up to 50 MB. For PDFs that is comfortably 100+ pages of typical scanned content.
How private is this?
Files travel over HTTPS to our hosted OCR + LLM stack, get processed, and are deleted within 24 hours. We do not train models on your uploads.