Clean up OCR text with AI

How it works

1. 1

   Extract the text from your scan with PDF OCR and PDF to Text.

2. 2

   Sign in and paste the OCR text.

3. 3

   Run the cleanup.

4. 4

   Copy the corrected text and proofread names and numbers.

About the PDF OCR Text Cleanup

Text recognised from a scan is rarely clean. You get "rn" read as "m", words split by hyphens and stray line breaks, and a paragraph cut into dozens of short lines. This tool takes OCR text you paste in and asks an AI model to fix those errors while keeping the meaning, paragraphs, lists and tables as they were. It returns only the corrected text, with no commentary.

What you can do

• Paste up to 20,000 characters of raw OCR output, about eight to ten pages of typical text.
• Fix misread characters, broken words and line breaks in the middle of sentences.
• Keep the paragraph structure, lists and simple tables of the original.
• Use any connected AI provider, since this tool works with plain text rather than a PDF file.

When it works best

• Tidying text from PDF OCR or PDF to Text before pasting it into a document.
• Cleaning up digitised letters, old typewritten reports or book scans.
• Preparing scanned text for translation, quoting or search indexing.

Things to know

• There is no file upload: get the text first with PDF OCR and PDF to Text, then paste it here.
• You must be signed in with a verified email, and pay with credits or an API key from Anthropic, OpenAI, Google or OpenRouter.
• The model corrects text it can reasonably infer. Words the scan destroyed completely may be guessed, so proofread names and numbers.
• Your text is sent to the chosen AI provider. Remove anything confidential first if that is a concern.

Frequently asked questions