Free OCR PDF to Word: Making a Page Allowance Go Further

A free OCR PDF to Word converter turns a scanned PDF into an editable Word file, up to a page allowance. To make the allowance cover what you need, convert only the pages you will work on, extract them into a small PDF before uploading, and join the Word pieces afterwards if the job takes more than one conversion.

The allowance is a budget, and a scanned document may not need all of its pages converted. A fourteen-page agreement may contain three pages you will edit and eleven you will only read. Spend the free pages on the three. The rest of this post is about choosing them, cutting the file to fit, and deciding when free has stopped being the cheap option.

What does a free OCR PDF to Word conversion hand back?

A free OCR PDF to Word conversion hands back a Word file holding the recognized text of the pages the allowance covered, and nothing for the pages it did not. Attached to it are the terms of the free tier: how many pages, how many conversions in a period, and how long the result stays available.

Before the first upload, write the allowance down. Four lines are enough.

The post on free scanned PDF to Word tools explains what free tiers limit in more kinds than these four, and sets out the Schema Law free plan in full. In brief, that plan allows one document a day of up to three pages and includes Word among its export formats.

What the Word file itself is

The file is recognized text: a typed reading of the page pictures, with the errors any reading has. Free or paid, it needs checking against the scan. The first check is arithmetic. Count the pages you uploaded and the pages that came back. A converter that stops at its cap without saying so gives you a file that looks finished and is not.

Whether the document should go to a free converter at all

A free converter is still someone else's computer. For a client document, read the post on whether free online OCR is safe for legal documents before uploading, and follow your firm's policy on online tools. The mechanics of the upload and download are covered in the walkthrough of an online OCR PDF to Word conversion.

Which pages of a scanned PDF deserve the free page allowance?

The pages that deserve a free allowance are the ones you will edit or quote at length, where retyping would cost the most time. Pages you will only read, pages with little text, and pages you already hold in another form do not need converting.

Go through the scan once and sort every page into one of four groups.

Pages you may already have

Before spending allowance on a page, ask whether its text exists somewhere else. Standard terms may be on the counterparty's website as a text PDF. A statute or court form reproduced in an exhibit is published by its source. Your own firm's precedent may be the document the scan was made from. Text from the source is exact. Recognized text is a reading of a picture and has to be checked.

A worked example

Lorne Castellan, a sole practitioner, holds a 14-page scanned services agreement between Brightwater Catering Ltd and the Hadleigh Conference Trust. The client wants the fee clause, the term clause and the termination clause changed. Castellan marks the scan:

• Pages 1 and 2: parties, recitals, definitions. He will read them and quote two definitions, which is quicker to type than to convert.
• Page 5: fees. Change.
• Page 9: term. Change.
• Page 10: termination. Change.
• Pages 13 and 14: signatures and a schedule of site addresses. No change.

Three pages hold everything he will edit. On a free tier capped at three pages a document, that is one conversion. The amendment he drafts will restate those three clauses and refer to the rest of the agreement by clause number, so the other eleven pages never need to be Word text.

Read the consideration on the fee page digit by digit before you draft. If the term clause states an option or a condition precedent, convert that whole sentence even when it crosses a page. The termination page earns the last free page because that is the clause you will rewrite.

Check the neighbours

A clause that begins at the foot of one page and ends at the head of the next needs both pages. So does a clause that leans on a definition or a schedule you will also have to change. Read the first and last lines of each chosen page on the scan before you settle the list, and add a page where a sentence runs over.

Cutting a scanned PDF down to a free converter's page cap

To cut a scanned PDF down to a free converter's page cap, extract the chosen pages into a new PDF with your PDF program and upload the extract. Extracting copies the page pictures unchanged. Name the new file with the original page numbers so you can map every Word page back to the paper.

Extracting the pages

1. Open the scan in your PDF program and find its page tools. The command is named Extract, Split or Organize Pages, depending on the program.
2. Select the pages on your list. If the program takes only a continuous range, extract each run separately.
3. Save the extract as a new file. Leave the original scan untouched.
4. Keep the pages in their original order inside the extract. Page 5, then 9, then 10.
5. Name the file with its source pages, such as Brightwater agreement pp 5, 9-10.pdf.
6. Open the extract and check it. Confirm the page count is at or under the cap and that each page is the one you meant.
7. Check the size. A converter may cap megabytes as well as pages.

Why extract and not print

Printing selected pages to a new PDF also produces a small file, and it is tempting when the PDF program on your machine has no extract command. Printing redraws each page, and depending on the printer settings the page pictures may be resampled at a lower quality than the scan. Recognition works from those pictures. Extract when you can, and if you must print, compare a printed page with the original at high zoom before uploading.

The page map

Once the extract exists, its page 1 is not the document's page 1. Write the map at the top of the Word file as soon as you open it:

Converted from scan of executed agreement. Word p. 1 = scan p. 5; Word p. 2 = scan p. 9; Word p. 3 = scan p. 10.

The map prevents a wrong page citation in whatever you draft from the file.

A PDF that will not let you extract

If the PDF program refuses to extract pages because the file is restricted, the restriction was set by whoever created the file. Ask the sender for an unrestricted copy. The post on converting a protected PDF to Word describes the kinds of protection and who can lift them.

Spreading a longer scanned document across several free conversions

A document longer than the page cap can go through a free converter in pieces, one conversion at a time, if the free tier's terms allow it and your deadline does. Divide the pages by the cap to get the number of conversions, divide by the daily limit to get the days, and join the Word files at the end.

The arithmetic

Take the pages you need, not the pages in the scan.

• Conversions needed = pages needed ÷ pages per document, rounded up.
• Days needed = conversions needed ÷ documents per day, rounded up.

On Schema Law's free plan, at three pages and one document a day, twelve pages are four conversions over four days. Thirty pages are ten days. Set the answer beside your deadline before you start. If the days needed are more than the days you have, the free route is closed for this document, however little the pages cost.

Read the terms of the free tier as well. A free allowance is offered on the provider's conditions. Use the one account you are entitled to, and do not open extra accounts to multiply the allowance.

A schedule you can keep

Put the pages you need soonest in the first piece. Prepare all the extracts on the first day, so each later day is an upload and a download and nothing more. Download each Word file on the day it is made. A free tier may remove files after a set time, so a ten-day job could lose its first pieces before the last was done if you left them on the service.

Joining the Word pieces

When the last piece is down, assemble one document with Word's Insert, Object, Text from File command, which Microsoft documents under inserting a document in Word. The post on large scans gives the full method for splitting a long scan by page range and rejoining the Word files, including reading across each join against the scan.

Pieces made from hand-picked pages on different days need three things more.

• Assemble in the document's page order, which may not be the order you converted in. The schedule put pages 5, 9 and 10 first because they were needed soonest. In the joined file they follow pages 1 to 4.
• Mark each join with a bracketed note of the scan page that starts there, and leave the notes in until you have proofread the whole file. Where the pieces skip pages you never converted, the note is the only sign of the gap.
• Make the styles consistent. Each piece was converted separately, so check heading levels and numbering across the whole file.

Is the Word file from a free OCR converter marked or restricted?

A Word file from a free converter may carry a watermark, arrive in a format other than DOCX, open read-only, or come with a download link that expires. Whether yours does is something to check in the file and in the converter's terms, not something to assume either way.

Six checks on the file you received

Terms that travel with the file

Some conditions are not visible in the document at all. Look in the converter's terms for three things: whether the free tier may be used for commercial or professional work, how long the download remains available, and whether the provider claims any right over uploaded content. If the terms restrict the output in a way that does not suit the matter, the answer is a different tool or a paid tier. It is not a way round the restriction.

Expiry

Treat every free result as temporary on the service. Download it when it is ready, open it, and save it to the matter folder. Then the service's retention period decides only how long your upload stays there, which is a confidentiality question. The post on whether OCR software stores your documents explains how to read a retention statement.

Retyping, a free conversion or one paid month for a single scanned document

For a single scanned document, retype when only a page or two of text is needed, use a free conversion when the pages fit the allowance and the deadline, and pay for a month when the page count or the calendar rules both out. The break-even is a calculation in your own time, not a fixed number.

Work out your three costs

Retyping. Time yourself typing one representative page from the scan, including proofreading it. Multiply by the pages you need. That is the cost in hours.

Free conversion. The money cost is nil. The time cost is the extraction, the uploads, the joining and the proofreading, plus the days the daily limit imposes. Proofreading is not optional on either converted route.

A paid month. The cost is the subscription price plus the proofreading. Schema Law's Solo plan is $49 a month for 1,000 pages, with up to 250 pages in one document, as the pricing page shows. Before counting it as a single month's expense, read how the subscription renews and ends. The post on legal OCR subscription terms covers what to look for, and the post on OCR cost per page does the per-page arithmetic.

A worked comparison

Suppose you need 36 pages of a scanned agreement as Word text, and you timed one page at eleven minutes to type and proofread.

Set six and a half hours of your own time, at whatever you count an hour as worth, against the month's price, and set twelve days against your deadline. Change the numbers and the answer changes. At three pages, the free tier wins outright: one upload, same day, no cost. At a single paragraph, retyping wins, because typing it is quicker than extracting a page.

A rule of thumb to test against your own figures

• Up to the free cap, no rush: free conversion.
• A few lines or one short page: retype.
• More pages than the free tier gives you before the deadline: a paid month, or a paid tool the firm already owns.
• More scanned documents expected next month: stop treating it as a one-off and price it as regular work.

Whichever route you take, the checking is the same. Read names, dates and amounts in the Word file against the scan before anyone relies on them.