How to convert image to text comes down to one technology: optical character recognition, or OCR, software that reads the pixels in a photo or scan and turns the letters it recognizes into selectable, editable text. The fastest free routes are the OCR already built into your phone and computer, Google Drive paired with Google Docs, and Google Lens. None of them is this compressor, and this guide walks through every one. For the other conversions this site covers, see the image conversion hub.
The catch nobody mentions is that OCR reads pixels and guesses at characters, so the quality of your source image decides how good the guess is. Below are the free methods, the two that actually get the text into Word and Excel, an honest answer about handwriting, and the prep that stops a good tool from producing garbage.
Does our tool convert an image to text?
No, and OCR was never part of the plan. This tool compresses and converts between four raster formats, JPG, PNG, WebP, and AVIF, and reading pixels to recognize the letters they form is a different kind of software, closer to what Google Docs and Adobe Acrobat run than anything a compressor does. If a page promises you a compressor that also does OCR, be skeptical of both features.
Where this tool genuinely helps is getting a scan small enough to OCR in the first place. Google Drive's OCR caps uploads at 2 MB, and a phone photo of a printed page straight off the camera routinely runs 4 to 6 MB before you touch it. Run one through the compressor here and the typical result, a 4.5 MB photo down to about 0.9 MB at quality 80, lands well under that 2 MB ceiling. One catch: Drive's OCR only reads .jpeg, .png, and .gif files, not WebP or AVIF, so choose JPG or PNG as the output format when you compress a scan for OCR.
What is OCR, and what can it actually do?
OCR is the technology that "converts scanned images of text into machine-readable text that you can search, copy, edit, and highlight" (Adobe Acrobat). It does not understand the words. It matches shapes in the image against letterforms it knows and outputs its best guess, which is why a clean scan of printed Arial converts almost perfectly and a dim photo of cursive comes back as noise.
Two distinctions matter before you pick a method. First, searchable is not the same as editable: some tools lay an invisible text layer under the original picture so you can find a word with Ctrl+F, while others hand you loose text you can rewrite. Second, a file that already contains real text is not an OCR job at all. If your source is a PDF, try to select or highlight the text; if you can, it already has a text layer and you are done, and if you cannot, you are likely dealing with a scanned PDF file that OCR is built for. Photos, screenshots, and plain image files never carry a text layer, so they always go through one of the methods below.
How to convert image to text: every free method
There are three tiers of free method, and most people reach for the online converters when the OCR already sitting on their device would have been faster. Start with what is built in, drop to Google Drive for a full page you want to edit, and use Lens for grabbing a quick passage.
The OCR already built into your phone and computer
Every current phone and desktop has an image to text converter built in, no upload and no account required. This is the method the top-ranked online tools never mention, because it competes with them.
On an iPhone or iPad, open the Photos app, tap the image that contains text, then press and hold on the text to highlight it and tap Copy. Paste it anywhere. On Windows, install the free Microsoft PowerToys, then press Win + Shift + T, drag a box over the text on screen, and it is copied to your clipboard instantly, which also works on text inside a video or an app window, not only image files. On a Mac, open the image in the Preview app, drag your cursor across the text the way you would select text in a document, and press Cmd + C.
These built-in tools are the right first move for anything short: a receipt total, a serial number, a paragraph off a slide. They keep no layout and export nothing to a file, but for copying a passage they beat every website.
Convert image to text free with Google Drive and Google Docs
For a full page you want to edit, Google Drive plus Google Docs is the strongest free route, and it costs nothing if you already have a Google account.
- Prepare the file: a PDF, or a
.jpeg,.png, or.gif, 2 MB or smaller, with text at least 10 pixels high, right side up, in a common font like Arial or Times New Roman, and shot with even lighting and clear contrast (Google Drive Help). - Upload the file to Google Drive.
- Right-click it and choose Open with, then Google Docs.
- Drive converts the image and opens a new Doc with the original picture at the top and the recognized text below it. Delete the image and keep the text.
The trade is layout. Google's own documentation says bold, italics, font size, font type, and line breaks are likely to survive, but "lists, tables, columns, footnotes, and endnotes are not likely to be detected" (Google Drive Help). Budget time to rebuild anything grid-shaped by hand.
Adobe Acrobat for scanned PDFs
Acrobat is the pick when your source is a scanned PDF you want to keep as a PDF, either searchable or fully editable. Online, for a PDF, it is free: drag the file into Acrobat's OCR tool, let it detect and convert the pages, then sign in to download the searchable result. On the desktop app, open the scanned PDF and choose All tools, then Scan and OCR, then Recognize Text, and Acrobat builds a searchable text layer under the original image so the page looks identical but the words are now selectable (Adobe Help Center). The limit worth knowing: the free online tool makes PDFs searchable, but turning an image format like PNG or JPG into editable text needs a seven-day free trial of Acrobat Pro.
Google Lens for a quick grab
Google Lens is the quickest way to pull a short passage out of a photo with no upload step. On desktop in Chrome, right-click the image, choose Search image with Google Lens, highlight the text in the result panel, and click Copy. On Android or iOS, open the photo in the Google or Lens app, tap the text option, highlight what you need, and tap Copy text (Google Lens). It copies plain text only and keeps no layout, so treat it as a clipboard tool, not a document converter.
| Method | Cost | Best for | Keeps layout? |
|---|---|---|---|
| Built-in device OCR | Free | Copying a passage fast, no file | No |
| Google Drive + Docs | Free | A full page you want to edit | Partly, text styling yes, tables no |
| Adobe Acrobat | Free for PDF search, Pro trial for images | Scanned PDFs, searchable or editable | Yes for the PDF layer |
| Google Lens | Free | Grabbing one paragraph or a number | No |
How do you get the extracted text into Word or Excel?
Getting the text into Word is a straight OCR job; getting a table into Excel is a harder problem the popular guides skip. Once you have editable text from any method above, Word is a paste away, but a spreadsheet needs the rows and columns preserved, and standard OCR flattens them.
For Word, the cleanest path is Google Drive plus Docs, then File, then Download, then Microsoft Word, which hands you a .docx with the styling that survived conversion. If your source is a PDF, Acrobat's Edit PDF applies OCR and exports to Word directly. Expect bold, italics, and line breaks to carry; expect to rebuild columns and footnotes.
Excel is the honest gap. General image to text converters read a table as a stream of words and lose the grid, because, Google states plainly that tables are not likely to be detected. To move a photographed table into a spreadsheet, you have two realistic options: use a tool that specifically does table OCR (Microsoft Excel's own Data from Picture feature in the mobile app is built for exactly this), or OCR the text with any method above, paste it into a sheet, and split it into columns with Data, then Text to Columns. Do not expect a generic photo to text converter to reconstruct a spreadsheet; that is a different job with a different tool.
| Destination | Best free route | Watch out for |
|---|---|---|
| Microsoft Word | Google Docs, then Download as Word | Tables and columns need rebuilding |
| Microsoft Excel | Excel mobile Data from Picture, or paste and Text to Columns | Generic OCR loses the grid entirely |
Can you extract text from an image of handwriting?
Sometimes, and far less reliably than printed text, so set expectations before you rely on it. To extract text from an image of neat, evenly spaced handwriting on a clean background, the built-in tools and Google Lens will often get most of it, and specialist handwriting engines like Transkribus do noticeably better on messy or historical script. But no free tool reads a shopping-list scrawl the way it reads a printed page.
The reason is the same one that governs every method here: OCR matches shapes against known letterforms, and handwriting varies far more than a typeface does. Cursive, slanted baselines, and joined letters all push recognition down. Treat any handwriting result as a draft and proofread every line, especially names, numbers, and anything you cannot verify from context. If accuracy matters and the volume is small, retyping is often faster than correcting a bad OCR pass.
How to get the most accurate OCR results
Accuracy is set by the source image, not the tool, so fix the image before you blame the software. Every method here reads pixels, which means a sharp, straight, well-lit capture converts cleanly and a dim or angled one does not, no matter which engine runs it.
Shoot or scan with even, flat lighting and a straight-on angle, since a shadow across half the page or text photographed at a slant is the most common cause of garbled output. Keep text at least 10 pixels high in the file you upload, which is Google's own floor, because anything smaller reads as noise. Stick to common fonts where you can, and crop out anything that is not text, since a table edge or a coffee ring gets read as characters.
File size is the quiet failure. Google Drive's OCR rejects anything over 2 MB, and phone scans routinely exceed that, so compressing first is often the difference between a clean pass and an upload error. The chart below is a real result from this tool: a 4.5 MB scan compressed at quality 80 lands at about 0.9 MB, comfortably under the limit, with no visible loss to the text.
| Stage | File size | Under Drive's 2 MB OCR limit? |
|---|---|---|
| Original phone scan | 4.5 MB | No |
| Drive OCR upload limit | 2.0 MB | The ceiling |
| Compressed at quality 80 | 0.9 MB | Yes |
Common mistakes that wreck OCR accuracy
Most bad results trace to a handful of avoidable errors, and knowing them saves a second attempt. Uploading a blurry or dim photo is the top one, because OCR cannot recover detail the camera never captured, so a fast retake beats cleaning up a bad shot. Expecting layout to survive is the second: running text converts well, but lists, tables, columns, and footnotes usually do not, so plan to rebuild them.
Confusing searchable with editable trips people up on Acrobat, where the free online tool makes a scanned PDF findable but editing an image format needs the Pro trial. Hitting Drive's 2 MB limit is common and easy to fix by compressing first. Skipping the proofread is the last and worst, since even a good pass carries occasional errors, so check names, numbers, and decimals before you trust the output. An online image to text converter is a starting point, not a final answer.