Self-hosted OCR · keyless · nothing stored

Turn any scan, photo, or PDF into text.

Drop a document and get clean, selectable text in seconds — on your own box. Sensitive invoices, contracts, and IDs never leave the building.

readwright · ocr language
Drop a scan, photo, or PDF
or click to choose · PNG, JPG, PDF · up to 20 MB / 15 pages · read then deleted
Reading your document… recognising the text on each page.
Selectable text out Private — never leaves the box Images + PDFs, multi-page 6 languages built in

Optical character recognition, without the cloud

Scanned paperwork, phone photos of receipts, multi-page PDFs. Readwright reads them into text you can copy, search, and paste.

1

Drop the file

An image or a PDF. Pick the language if it isn't English. We rasterise each page and read it.

2

See it side by side

Your document on the left, the extracted text on the right — page by page, fully selectable.

3

Copy or download

Grab the plain text or download a .txt. Nothing is stored after the request.

For the paperwork you can't paste into a random website

The documents worth digitising are usually the ones you'd never upload to a free cloud OCR.

🧾

Finance & ops

Invoices, receipts, and statements into text for bookkeeping and expense systems — no per-page meter.

📄

Legal & admin

Contracts, forms, and scanned records digitised on your own infrastructure, not a third party's.

📚

Research & archives

Books, articles, and historical scans turned into searchable text in bulk.

Pricing

Read documents free in the browser. Upgrade for longer files, more languages, and an API.

Free

$0
  • Up to 15 pages
  • Images + PDFs
  • TXT export
  • Nothing stored
Read a document
Popular

Creator

$9/mo
  • Long PDFs
  • Batch upload
  • All languages
  • Priority queue
Start Creator

Pro / API

$29/mo
  • REST endpoint + key
  • Bulk OCR
  • Searchable-PDF output
  • Webhooks
Get API access

Self-host

$249once
  • Run on your infra
  • Unlimited pages
  • Air-gap friendly
  • Updates for 1 year
Get license

Questions

Do you keep my documents? +

No. Each file is written to a temp path, read, and deleted in the same request. There's no database and nothing is retained.

How accurate is it? +

It uses the open Tesseract engine, which is strong on clear printed text. Clean scans and sharp photos read best; low-resolution or handwritten pages are harder.

Which languages? +

English, Spanish, French, German, Portuguese, and Italian are built in; the self-host license adds any of Tesseract's 100+ language packs.

Can I run it on my own servers? +

Yes — the self-host license runs the whole OCR service on your infrastructure, air-gap friendly, with no per-page cost.