PDF to Text and Images

You have a PDF: a contract, a report, an ebook, a slide deck someone sent you, a scanned form. You want the words out of it, or you want an AI to read it. PDF to Text and Images does both. Drop the PDF in and you get clean, readable text you can copy or download, page images you can hand to an AI, and one-click prompts that send the document to the evyAI Agent.

It lives at evyus.com/tools/pdf-to-text, it is free, needs no login, and runs in your browser: the PDF is never uploaded to a server.

Who it’s for

Anyone who needs the text of a PDF to quote, edit, or paste somewhere else.
Anyone who wants an AI to summarize, explain or push back on a document.
Anyone stuck with a scanned PDF where you can’t select the text.
Content creators turning an ebook, a report or a deck into posts.

How it works

1. Drop in your PDFs

Drag one or more PDFs onto the page, paste one, or click to choose. Each file gets its own tab at the top so you can switch between them. If a PDF is password protected, you are asked for the password, and it stays in that browser tab.

Once a file is read you see the number of pages, words, characters, an estimate of AI tokens (how much of an AI’s reading limit the document uses) and the reading time.

2. Pick the pages you want

Every page has a thumbnail on the left. Click a thumbnail to jump to that page’s text. Tick the box under a thumbnail, or type a range in Pages (for example 2-5, 9), to work with only some pages. That choice applies to everything: the text, the page images and what you send to evyAI. Leave it empty for the whole document.

3. Get the text

The Text tab shows the text page by page. Choose Plain text or Markdown (Markdown keeps headings and bullet lists as formatting, which AI tools read well). Four switches control the cleanup:

Join lines into paragraphs: turns the PDF’s short lines back into normal paragraphs.
Remove headers, footers and page numbers: drops the lines that repeat at the top or bottom of most pages.
Fix words split by a hyphen: rejoins words broken across two lines.
Show page breaks: marks where each page starts.

Tables come out as rows with the cells split by |, so they still make sense when pasted. Use Find in text to search, and Copy page to copy a single page.

Send to Post Editor opens the Post Editor in a new tab with the text of the pages you picked, ready to cut down, style and preview as a LinkedIn or X post. Anything you were already writing there is saved to My drafts first.

Send to Carousel Maker opens the Carousel Maker in a new tab with the text of the pages you picked already loaded. Let evyAI turn it into slides, or split it into slides yourself; headings in the PDF become slide headings. A carousel works best from a few pages, so tick just the ones you want.

4. Read scanned pages with OCR

Some PDFs are just pictures of pages, so there is no text to pull out. When that happens a yellow banner tells you which pages have no text. Pick the language and press Read them with OCR. The pages are read one at a time right in your browser, and the text appears with a read by OCR tag. The first time, the browser downloads the OCR engine, which takes a few seconds.

5. Turn pages into images for an AI

The Page images (JPEG) tab turns pages into image files. This is the best way to have an AI read anything text alone loses: charts, tables, forms, handwriting, signatures and page layout. Choose:

Size: Best for AI chat (1568 px long side) is the default, about the largest size AI chat apps read without shrinking it. There are sharper and lighter sizes, plus 150 and 300 DPI for print or archive copies.
Format: JPEG, PNG (lossless) or WebP (smallest), and the quality.
Pages per image: 1, 2 side by side, or 4 in a grid. Each page is labeled with its number, so the AI can tell you which page it means. This helps when your chat app limits how many images you can attach.
Black and white for smaller files.

Press Make images. Each image has Download and Copy (to paste straight into a chat). Download all (.zip) gets every image, and Images + text (.zip) puts the images and the Markdown text in one file. Then drag the images into evyAI, ChatGPT, Claude or Gemini with your question.

6. Send it to evyAI

The Send to evyAI tab has one-click prompts that open the evyAI Agent with the document already in the chat:

Summarize it, Key takeaways, Action items, Explain it simply
Questions to ask before you sign, buy or meet, and Push back on it for gaps and risks
LinkedIn post, Carousel slides (written in the format the Carousel Maker reads), Email summary, Best lines, Checklist or SOP and FAQ

Or type your own instruction and press Send to evyAI. Long documents are too big for a link, so the prompt goes on your clipboard and evyAI opens for you to paste it; the page tells you before you click. Copy the prompt instead lets you use any other AI.

7. Details and links

The Details and links tab shows the title, author, dates, page size and whether the file is locked. If the PDF has bookmarks you get a clickable contents list, and every link inside the PDF is listed with its page. Copy the links, download them as CSV, or open them all with the Bulk Link Opener.

What comes out

Copy all, or download .txt, .md or .json (every page separately, for developers). With two or more PDFs open, All files .zip downloads the text of each one.
Page images as JPEG, PNG or WebP, one at a time or in a ZIP.
Summaries, posts, checklists and answers written by evyAI.
A Recent list keeps the text of your last 8 PDFs in this browser, so you can come back to it without the file.

Where it fits

PDF to Text and Images is where a document enters the content loop on the evyAI Tools overview. Take the best idea to the Hook Lab for an opening line, polish the post in the Post Editor, or turn the carousel prompt’s answer into slides in the Carousel Maker. For audio and video instead of PDFs, use Voice to Text.

Limits

Text is pulled from what the PDF stores. Unusual layouts (many columns, text inside images, complex tables) may come out in an odd order. When the layout matters, use page images instead.
OCR is good, not perfect. Check names, numbers and dates against the page. It reads 11 languages, one language at a time.
OCR runs only on pages that have no text at all, not on pictures inside an otherwise normal page.
Very large PDFs (hundreds of pages) can take a while and use a lot of memory, especially when making high resolution images.
The Recent list and your settings live in this browser on this computer. The page images are not saved; make them again or download the ZIP.
Something not working? Use Report a problem on What’s New.