Back to home

Documentation

Everything the app can do, and how to use it — from a single upload to bulk archives.

Getting Started

There are three ways to get a document into the extractor:

  • Upload a file — drag and drop, or click to browse. PDFs and PNG/JPG images are supported.
  • Take a photo — use your device's camera directly in the browser, with pinch-to-zoom, a flash/torch toggle, and crop & rotate before extracting.
  • Scan with Phone — from a desktop without a camera, hand the shot off to your phone (see below).

Pick an OCR language (or leave it on auto-detect) and an engine, then start extraction. Progress and an estimated time are shown live while it runs.

OCR Engines

Vision AI Pro is the recommended engine — an AI-powered recognition model tuned for the highest accuracy, especially on handwriting, photos, and non-Latin scripts. It's included on Starter and Pro plans, and available to anyone without a subscription via pay-as-you-go word credits.

Every plan, including Free, also includes a standard baseline extraction engine for straightforward documents. If a scan doesn't come out clean on the standard engine, switching to Vision AI Pro (or picking up a small word-credit pack) is the most reliable fix.

Elite Vision, a third engine for the most demanding documents, is in development and not yet available.

AI Enhancement (“AI Polish”)

Available on Starter and Pro plans. After extraction, AI Polish automatically re-reads the text to correct OCR errors, fix broken formatting, and tidy up spacing — without changing the meaning of the original document. You can always toggle back to the original, unedited extraction.

Fixing Mistakes by Hand

Select any word or phrase in your extracted text (right-click on desktop, or tap-and-hold on mobile) to open the fix menu, with three options:

  • AI fix — re-run correction on just that snippet.
  • Edit it myself — type a manual correction. For Nepali and Hindi, this supports phonetic typing: type in Roman letters and it converts to Devanagari as you go.
  • Rescan from image — crop the matching region of the original photo and re-run OCR on just that area, which often fixes a single misread word without redoing the whole page.

Edits are saved back into that scan's history entry automatically.

Language Support

30+ languages are supported, with automatic language detection and a per-page breakdown shown for documents that mix multiple languages (common with Nepali/Hindi/English multi-script documents).

Scan with Phone

Available on Starter and Pro plans. On a desktop without a good camera, click "Scan with Phone" to generate a QR code. Scan it with your phone, take the photo there, and it's relayed back to your desktop session automatically and staged for extraction — no app install, no file transfer required.

Bulk Processing

Available on the Pro plan. Upload many files at once and process them as a batch, with per-file progress and a single combined export when it's done — built for digitizing larger archives rather than one document at a time.

When the batch finishes, results are always available as a single ZIP download on any device, including phones. On desktop Chrome and Edge you can additionally pick an output folder before processing, and the finished files are saved straight into it.

Scan History

Every extraction is saved automatically so you can find it again later. History is searchable by file name, and retention depends on your plan — 7 days on Free, 30 on Starter, 90 on Pro. Reopen a past scan to view, re-edit, or delete it at any time.

Sharing a Scan

From any history entry, choose exactly what to share: the extracted text, the original scanned image, or both.

  • On mobile — uses your device's native share sheet, so you can send it straight to Messages, WhatsApp, Gmail, or anywhere else installed.
  • On desktop — downloads the file(s) directly, ready to attach wherever you need them.
  • Email it — type any recipient's address and we'll send it for you as a real email attachment, on any device.

The original image is retained for sharing on Starter/Pro plans; Free-tier scans keep the extracted text only.

Account & Syncing Devices

Plans and word credits are tied to your email, not a single browser. To use a paid plan or credit balance on a new device, go to Account, enter the email you used at checkout, and we'll email a one-time code. Entering it remembers that device for 30 days — no password to remember, and nobody else can sync your account without access to your inbox.

Plans & Word Credits

Subscriptions (Free / Starter / Pro) unlock higher limits, faster engines, and extra features — see the full breakdown on Pricing.

Prefer not to subscribe? Word credits are a one-time purchase, metered per word, redeemable against Vision AI Pro — no recurring charge, and credits never expire.

Output Formats

TXT and PDF are available on every plan. Starter adds DOCX, JSON, and Markdown. Pro adds CSV and XML for structured/batch workflows.

Privacy & Support

Read the Privacy Policy and Terms of Service for details on data handling. For anything else, reach out any time at contact@aetherai.store.