How I Convert Images to Text on My iPhone (And the Tools I Actually Use)

I used to retype an entire page from a textbook because I couldn’t copy the PDF. It took twenty minutes and I still had typos. That was the moment I decided to figure out how to convert images to text on my iPhone without losing my mind—or my afternoon.

Here’s what I actually use now after two years of trying different approaches. No fluff, just the three methods that survived into my daily routine and when each one genuinely earns its spot.

The moment I realized I needed this

It started with a printed handout from a meeting: names, phone numbers, action items, a couple of deadlines. I needed that info in a spreadsheet. My options were typing it all out—half an hour, guaranteed errors—or figuring out a faster way. I knew phones could “read” text from photos; I’d seen the feature mentioned online a dozen times. But I’d never actually built a workflow around it. That handout was the push.

A week later I found myself doing it again with a restaurant receipt. Then a whiteboard photo from a brainstorm session. Then my handwritten notes from a conference. The pattern was clear: I needed a reliable way to go from photo to editable text on my iPhone, and I needed it to work for different kinds of images.

I tried a few things, messed up a few times (including the time I accidentally uploaded a receipt with my card number to a random website—more on that later), and eventually landed on a simple decision tree. If you’ve ever stared at a photo of a menu or a whiteboard and wished it was just text, this is for you.

Three ways I convert images to text on iPhone, ranked

I don’t use one tool for everything. That was my first mistake—assuming there’d be a single perfect option. Instead, I pick based on three factors: how much text there is, whether it’s printed or handwritten, and whether I need to export it somewhere or just copy a line.

1. A dedicated OCR app — when I need accuracy, handwriting, or exports

When the photo has a lot of text, messy handwriting, or I need a clean file I can paste into Notion or send to someone, I reach for an OCR app. I’ve been using Textora because it runs on-device (my images stay on my phone), handles both printed and handwritten text, and gives me export options. But the principle applies to any serious OCR app: you feed it an image, it gives you structured text back.

The tradeoff is workflow friction. I have to open the app, pick or take a photo, wait a second for processing, then copy or export. For a single sentence on a sign, that’s overkill. For a page of meeting notes or a stack of receipts I need to file, the accuracy and export capability are worth those extra taps.

What surprised me: handwriting recognition has gotten dramatically better in the past year. I write in a mix of print and cursive that even I have trouble reading sometimes, and modern on-device OCR handles about 85-90% of it correctly. Two years ago that number was closer to 60%.

2. iOS Live Text — when I just need to copy a quick line

If I’m already in Photos or Safari and I only need one line or a short paragraph, Live Text is unbeatable. Tap and hold the text, hit Copy, done. No app to open, no uploading, no waiting. It’s baked into the system and works in more places than most people realize: the Camera app, Quick Look, even some third-party apps.

The feature that really sold me: you can point the Camera at text without taking a photo and copy directly from the viewfinder. I use this at coffee shops to grab WiFi passwords off signs, and at pharmacies to copy the drug name off a box so I can look it up. It feels like a superpower the first few times you use it.

The catch is that Live Text has a ceiling. Long blocks of text? It often misses lines or rearranges paragraphs. Handwriting? Sometimes it doesn’t even offer the option to select. Complex layouts like a newspaper column or a receipt with columns? It can scramble the order. For “grab that one line,” it’s perfect. For anything more, you’ll outgrow it quickly.

3. Google Lens — when I also need translation

When the text is in another language and I want to understand it—not just copy it—I use Google Lens. Point the camera or pick a photo, and it overlays a translation or gives you a copyable version. I relied on this heavily during a trip to Istanbul last year: menus, signs, even instructions on a washing machine at the Airbnb.

For English-only text or when I need a clean export, I stick with the other two methods. Lens is the tool I pull out for “what does this say?” in a language I don’t speak. It’s also decent at identifying things in photos (plants, products, landmarks), which is a nice bonus when you need it.

The limitation: Lens wants to be connected to the internet for full features, especially translation. Offline mode exists but it’s limited. And it’s not designed around “give me a clean block of text I can export”—it’s more of a quick-answer tool.

Mistakes I made so you don’t have to

Mistake 1: Using the wrong tool for the job. I spent ten minutes trying to get Live Text to read a full page of handwritten meeting notes. It kept selecting random fragments and ignoring half the lines. When I finally switched to an OCR app, I had clean text in fifteen seconds. Match the tool to the task.

Mistake 2: Not cropping first. This one cost me a lot of wasted time. When the photo has a lot of background—your desk, your hands, a coffee cup—the OCR engine has to figure out what’s text and what isn’t. Cropping so the text fills most of the frame improved my results dramatically. Five seconds of cropping saves five minutes of fixing gibberish.

Mistake 3: Shooting at an angle. I used to just snap a photo from whatever angle was convenient. Turns out OCR engines are optimized for straight-on shots. Angled text gets geometrically distorted, and characters that look fine to your eyes get misread by the software. Now I take an extra second to square up, and my error rate dropped noticeably.

Mistake 4: Ignoring the privacy angle. Early on I used a random online OCR website to convert a receipt. It worked fine—but the receipt had my name, the last four digits of my card, and the amount. I have no idea where that image ended up. Now I only use on-device tools for anything remotely sensitive. If the image doesn’t leave my phone, I don’t have to worry about it.

When each tool fails and what to do about it

Every method has a ceiling, and knowing where that ceiling is saves you from wasting time with the wrong approach.

• Live Text fails on handwriting, decorative or unusual fonts, very small text, and complex multi-column layouts. Sometimes it simply doesn’t surface the copy option at all. Fix: switch to a dedicated OCR app that’s built for documents and handles layout analysis.

• OCR apps can still struggle on blurry or low-contrast images—pencil on yellow legal paper, faded thermal receipts, or photos taken in dim lighting. Fix: retake the photo with better light and a steady hand, increase the contrast in the Photos editor, or crop tighter to the text area. A clean image is worth more than a better algorithm.

• Google Lens needs a data connection for translation, can be slow on long text, and sometimes mangles the order of dense paragraphs. Fix: for long, non-English text, I sometimes take a screenshot of the translation and run that through an OCR app to get clean, copyable text.

The takeaway: when one method gives you garbage, try the next one on the list before assuming the text is unreadable. I’ve had images that Live Text couldn’t touch but an OCR app handled perfectly, and vice versa.

My actual daily workflow (with real examples)

For a restaurant menu I want to share in a group chat: I open the Camera, point at the menu, use Live Text to copy the dish names, and paste into Messages. No photo saved, no app opened. Total time: ten seconds.

For a whiteboard after a brainstorm: I take a photo, make sure it’s straight and well-lit, open an OCR app, extract the full text, and paste into Notion with the heading “Brainstorm Notes [date].” The OCR handles the mix of printed labels and handwritten scribbles. Total time: about thirty seconds.

For screenshots (error messages, tweets, recipe cards from Instagram stories): if it’s one line, Live Text right from the screenshot in my Photos library. If it’s a whole thread or a long error log, I run the screenshot through an OCR app so I can copy the entire text block at once. I wrote more about copying text from screenshots in a separate post—there are some tricks with Shortcuts that can speed this up even more.

For handwritten notes from a lecture or meeting: an OCR app is the only reliable option. Live Text isn’t consistent enough for handwriting that isn’t perfectly neat. I make sure the page is well-lit, I’m shooting straight down, and the writing fills the frame. Under those conditions, accuracy is surprisingly good. Under bad conditions (dim room, angled shot, tiny handwriting), even the best app will struggle.

For receipts and business cards: an OCR app, always. I need the text structured and exportable, not just a quick copy. For business cards specifically, the app can usually pull out the name, title, email, and phone number separately—which is much more useful than a blob of text.

Quick comparison

Tool	Best for	Handwriting	Export / long text	Offline
Live Text	Quick copy, one line	Weak	No	Yes
OCR App	Accuracy, batch, export	Strong	Yes	Usually
Google Lens	Translation, quick lookups	Okay	Limited	Partial

The bottom line

I don’t force one tool for everything. I match the tool to three variables: how much text, what kind of text, and where I need the result to end up. That decision takes two seconds and saves ten minutes of fighting the wrong method.

The order I check: Live Text first (it’s already there). If it can’t handle it, an OCR app. If I need translation, Google Lens. I haven’t found a situation that doesn’t fit one of those three.

If you want to go deeper on iPhone-only options (including where Live Text hides in places you might not expect), check out every way to copy text from an image on iPhone. For a straight-up app comparison with real tests on the same images, see photo-to-text apps compared.