# How can I extract text from pdf for my contract reports?

**URL:** <https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42>\
**Category:** Accessible PDFs\
**Tags:** outreach\
**Created:** [October 29, 2026, 3:14pm UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42 "2026-10-29T15:14:17Z")\
**Posts on this page:** 12\
**Page:** 1

<div class="post-metadata">

**Author:** ![NovaBailey](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/n/9f8e36/32.png) [@NovaBailey](https://socialgoodconnect.org/u/NovaBailey)\
**Post date:** [October 29, 2026, 3:14pm UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/1 "2026-10-29T15:14:17Z")

</div>

I need to pull specific clauses from several scanned contract PDFs for my quarterly report, but copying and pasting just scrambles the formatting. What is the best way to cleanly extract this text?

---

<div class="post-metadata">

**Author:** ![temp964610](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/t/aca169/32.png) [@temp964610](https://socialgoodconnect.org/u/temp964610)\
**Post date:** [October 29, 2026, 3:14pm UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/2 "2026-10-29T15:14:20Z")

</div>

Scanned pages are just images, so you need OCR first. Try opening the PDF in Google Drive with Google Docs, it pulls the text out surprisingly well. For batches, Acrobat’s OCR tool keeps layout better.

---

<div class="post-metadata">

**Author:** ![kind\_yeti](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/k/54ee81/32.png) [@kind\_yeti](https://socialgoodconnect.org/u/kind_yeti)\
**Post date:** [October 29, 2026, 5:21pm UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/3 "2026-10-29T17:21:16Z")

</div>

Scanned contracts are honestly the worst for this. It is so frustrating when copied text turns into complete gibberish! Since they are just pictures of paper, you need a program that actually recognizes the letters. I usually run my scanned paperwork through Adobe Acrobat, or upload it to Google Drive and open as a Google Doc to grab the clean text.

---

<div class="post-metadata">

**Author:** ![samanthag](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/s/34f0e0/32.png) [@samanthag](https://socialgoodconnect.org/u/samanthag)\
**Post date:** [October 30, 2026, 11:36am UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/4 "2026-10-30T11:36:03Z")

</div>

I had this with scanned leases, and OneNote’s “Copy Text from Picture” gave me cleaner text than copying from the PDF. Paste each clause into Word, then fix headings and line breaks there. Check names, dates and amounts against the scan because text recognition can make mistakes.

---

<div class="post-metadata">

**Author:** ![vinyl\_fan](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/v/db5fbb/32.png) [@vinyl\_fan](https://socialgoodconnect.org/u/vinyl_fan)\
**Post date:** [October 30, 2026, 12:21pm UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/5 "2026-10-30T12:21:05Z")

</div>

That can be messy, especially with scanned contracts. I would use Google Lens to scan each needed page, copy the recognized text into Word or Google Docs, then clean up the paragraph breaks and italics there. Check every clause against the scan before using it.

---

<div class="post-metadata">

**Author:** ![JasonDoyle](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/j/ac91a4/32.png) [@JasonDoyle](https://socialgoodconnect.org/u/JasonDoyle)\
**Post date:** [October 30, 2026, 12:36pm UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/6 "2026-10-30T12:36:05Z")

</div>

[vinyl\_fan](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/5) I do something similar with our intake scans. I run OCR first, paste only the needed clauses into Word, then use Show/Hide to remove odd line breaks. Definitely compare names, dates, and dollar amounts back to the original page.

---

<div class="post-metadata">

**Author:** ![xXRedXx](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/x/8e8cbc/32.png) [@xXRedXx](https://socialgoodconnect.org/u/xXRedXx)\
**Post date:** [October 30, 2026, 3:21pm UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/7 "2026-10-30T15:21:06Z")

</div>

If its scanned, regular copy paste wont work since theres no text layer, that’s your scrambling issue.

1. Run OCR first, I use Adobe scan or sometimes just Acrobat’s “recognize text” function.
2. After OCR, export as Word or plain txt instead of copy pasting straight from the PDF viewer.
3. Check the clauses after, OCR messes up numbering and bullet points sometimes.

---

<div class="post-metadata">

**Author:** ![sneaky\_grandpa](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/s/7cd45c/32.png) [@sneaky\_grandpa](https://socialgoodconnect.org/u/sneaky_grandpa)\
**Post date:** [October 31, 2026, 1:06am UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/8 "2026-10-31T01:06:48Z")

</div>

Scanned PDFs are just pictures, so copy/paste really is guesswork. Best way I’ve found is to run them through PDFcap, turn on OCR, then export to text; the clauses come out clean and in order. My old laptop grumbles on big batches, so I do ten pages at a time. If the scan is crooked, I straighten it first or the OCR gets silly.

---

<div class="post-metadata">

**Author:** ![xxx\_night\_xxx](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/x/db5fbb/32.png) [@xxx\_night\_xxx](https://socialgoodconnect.org/u/xxx_night_xxx)\
**Post date:** [October 31, 2026, 6:06pm UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/9 "2026-10-31T18:06:57Z")

</div>

[Vinyl Fan](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/4) I do the same with old lease contracts. Google Lens on the phone is easy, but for longer pages I find opening the recognized text in Word and turning on show/hide helps me spot the bad line breaks. Did yours come out cleaner on the phone or the desktop?

---

<div class="post-metadata">

**Author:** ![phoenix\_wind](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/p/ea666f/32.png) [@phoenix\_wind](https://socialgoodconnect.org/u/phoenix_wind)\
**Post date:** [November 1, 2026, 2:51pm UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/10 "2026-11-01T14:51:03Z")

</div>

For scanned PDFs, I’d OCR a copy first, then paste the needed clauses into Word and rebuild the formatting; OneNote’s “Copy Text from Picture” can work for a few pages. Tip: compare every quote with the scan, especially numbers and dates. Avoid uploading confidential contracts to random online services, tbh.

---

<div class="post-metadata">

**Author:** ![olivia\_bulls](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/o/a9adbd/32.png) [@olivia\_bulls](https://socialgoodconnect.org/u/olivia_bulls)\
**Post date:** [November 3, 2026, 12:36am UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/11 "2026-11-03T00:36:21Z")

</div>

I open the scanned PDFs in Word now and copy the clauses from there. Copying straight out of the PDF always scrambles the formatting for me. fwiw I learned that the hard way on last quarter’s report.

---

<div class="post-metadata">

**Author:** ![temp.8866](https://socialgoodconnect.org/letter_avatar_proxy/v4/letter/t/e47774/32.png) [@temp.8866](https://socialgoodconnect.org/u/temp.8866)\
**Post date:** [November 7, 2026, 2:36am UTC](https://socialgoodconnect.org/t/how-can-i-extract-text-from-pdf-for-my-contract-reports/42/12 "2026-11-07T02:36:24Z")

</div>

@olivia_bulls Opening them directly in Word has worked best for me too on office handouts. My old laptop slows down during the conversion, but it usually gets the words right.

What device are you running it on?
