MemX is live on Product Hunt today.Please support us(opens in a new tab)
AI Tools

Find Any Photo Just by Describing It

Arpit TripathiArpit TripathiLinkedIn·July 22, 2026·12 min read

Native photo search needs the right phone and only checks one gallery. Here is how to find any photo, screenshot or doc by describing what is in it.

Yes, you can now find a photo by describing what is in it, and the days of thumb-scrolling through thousands of thumbnails are ending. Type "receipt from the hardware store" or "whiteboard with the flowchart" into a modern phone gallery and it can surface the match, which is exactly what people mean when they ask how to find a photo by describing it or how to search photos by description.

The catch is that the good version is gated, and the line is drawn higher than most people expect. A brand-new standard iPhone 15 does not get Apple's natural-language photo search, and a Pixel 8 cannot run Google's screenshot search at all. Apple's version runs only on devices that support Apple Intelligence, Google's Pixel Screenshots search is fenced to a short list of phones, languages, and countries, and every one of these tools searches only its own gallery. None of them looks inside the screenshots you saved in a chat app, the PDFs in your files, or the scanned documents in your email, and none tells you which specific image actually answered your question. Below: how each native tool works, where it stops, and how one searchable index across everything you own closes the gap.

What "search photos by description" really means

Searching photos by description means the phone understands the contents of an image, not just its filename or date, so you can retrieve it with plain words. Two different technologies make this work. The first is OCR, or optical character recognition, which reads the actual text printed inside a photo: the numbers on a receipt, the words on a whiteboard, a Wi-Fi password on a router label. The second is semantic image understanding, which recognizes objects, scenes, and actions so a search for "dog on a beach" returns the right frame even though no text appears in it.

Basic versions of both have existed on phones for years. Apple's Live Text has read text inside photos since iOS 15 on iPhone XS and later, and the Photos app has long matched simple categories like beach, dog, or car on any recent iPhone. What changed recently is the jump to full natural-language phrases, where you can string together objects, colors, people, and settings in one query. That richer search is the part with hardware strings attached.

How to search iPhone photos by description

On iPhone, describing a photo in a full sentence works only if your device supports Apple Intelligence. Apple added natural-language search to Photos in iOS 18.1, letting you search phrases like "kids laughing at swings," "couple dancing at wedding," or "yellow flowers in vase by window," and it even reaches into videos with queries like "baby crawling." It runs only on iPhone 15 Pro, iPhone 15 Pro Max, and the entire iPhone 16 and iPhone 17 lineups. Older phones, including the standard iPhone 15, iPhone 14, and anything earlier, do not get it.

If your iPhone does not support Apple Intelligence, you are not blocked entirely. Simple keyword search for people, pets, common objects, and scenes still works, and Live Text still lets you search for words that appear inside a photo on iPhone XS and later. What you lose is the descriptive, multi-part phrasing that makes search feel conversational. In practice that means an iPhone 13 owner can search "dog" but not reliably "brown dog catching a frisbee at the park."

A separate feature, Enhanced Visual Search, handles landmarks. Introduced in iOS 18, it lets you search your library for places like the Golden Gate Bridge or the Arc de Triomphe, and it does not require Apple Intelligence. Your device computes an encrypted mathematical representation of a possible landmark, sends it to Apple to match against a global list using homomorphic encryption and other privacy techniques, and receives encrypted results back. Apple turns it on by default, it is not available in every country or region, and it is limited to landmarks rather than general "describe anything" search.

Insight

Two iPhone features get mixed up, and the difference decides what your phone can do. Natural-language phrase search needs Apple Intelligence, so an iPhone 15 Pro, iPhone 16, or iPhone 17. Enhanced Visual Search for landmarks runs on iOS 18 more broadly but only matches famous places and is region-limited. Neither one reads your PDFs, and neither searches screenshots stored outside the Photos app.

How to search Pixel screenshots by description

On Android, the closest match to "describe it and find it" is the Pixel Screenshots app, and its reach is narrow. The app uses on-device AI to generate titles and summaries for your screenshots and lets you search them with natural language, then tap suggested action chips. Google restricts it to Pixel 9, Pixel 9 Pro, Pixel 9 Pro XL, Pixel 9 Pro Fold, Pixel 10, Pixel 10 Pro, Pixel 10 Pro XL, and Pixel 10 Pro Fold. No other Pixel and no other Android phone qualifies.

The language and country limits are just as tight. The Pixel Screenshots app supports only English, German, and Japanese, and it is available only in Australia, Canada, Germany, India, Ireland, Japan, Malaysia, Singapore, the United Kingdom, and the United States. Google also notes that all AI features are disabled if you sign in with an ineligible account, and that the AI models download over Wi-Fi only. Miss any one of those conditions and the natural-language search simply is not there.

The name gives away a deeper limit: Pixel Screenshots searches screenshots. It is built around the images you capture with the power-plus-volume shortcut, not the photos in your main camera roll, not the PDF a school emailed you, and not a document you scanned. So even a Pixel 10 owner in a supported country still has their saved contracts, statements, and photographed notes sitting outside the tool's reach.

Why native photo search misses screenshots, PDFs, and documents

Every native tool searches one island and ignores the rest of your files. Apple Photos searches the Photos library. Pixel Screenshots searches screenshots. Neither reaches across to the other, and neither indexes the documents where a lot of important information actually lives. A single answer might hide anywhere: a screenshot of a boarding pass saved in a chat thread, a PDF lease in your email, a photo of a prescription label, a scanned warranty card in your files app. The thing you need could be in any of them, and no gallery search spans all four.

  • Screenshots you saved inside a messaging app or downloaded from the web often never touch the native gallery search index.
  • PDFs and office documents are invisible to phone photo search, even when they contain the exact text you are hunting for.
  • Voice notes and scanned paper live in yet another app, disconnected from any of it.
  • Search results rarely cite the source. You get a filtered grid of images and still have to eyeball which one holds the answer.

When you ask a person "what was the Wi-Fi password at the rental," you expect them to point at the exact sticky note. Native photo search hands you a wall of candidate thumbnails and leaves the final read to your eyes. It narrows the haystack; it does not hand you the needle with a receipt showing where it came from.

Native photo search compared

Of the three, only MemX searches across your gallery, screenshots, PDFs, and documents at once and returns the specific source image with the answer; Apple Photos and Pixel Screenshots each search a single walled library.

CapabilityApple Photos (iPhone)Pixel Screenshots (Android)MemX
Search by describing image contents in a full phraseYes, but requires Apple Intelligence (iPhone 15 Pro, 16, or 17)Yes, but only Pixel 9 and Pixel 10 modelsYes, on any phone through the app or WhatsApp
Reads text printed inside images (OCR)Yes, via Live Text on iPhone XS and laterYes, for screenshots on supported PixelsYes, across all photos, screenshots, and documents
Searches screenshots saved outside the galleryNoNo, screenshots app only, not other saved imagesYes, anything you add is indexed together
Searches PDFs and scanned documents tooNoNo, screenshots onlyYes
Returns the specific source image with a citationNo, shows a grid to scanShows the matching screenshotYes, answers in plain English and cites the source
Works across iPhone, Android, and WhatsAppNo, iPhone onlyNo, select Pixel onlyYes
Language and country limitsEnhanced Visual Search is region-limited; Apple Intelligence rolls out by language3 languages, 10 countriesAvailable on Android, iOS, and WhatsApp

How to find a saved Wi-Fi password or whiteboard photo

Two everyday moments show exactly where native search falls short. The first is the screenshot with the Wi-Fi password. You snapped a photo of the router sticker or screenshotted the credentials someone messaged you, then buried it under a thousand later pictures. To retrieve it you need OCR that read the letters and numbers on that label plus a search that spans wherever the image ended up. If the screenshot lives in a chat app rather than your camera roll, iPhone and Pixel gallery search may never see it.

The second is the whiteboard photo. After a meeting or study session, you photograph a whiteboard full of diagrams, action items, and a URL. Weeks later you want the one line about next steps. Basic gallery search might surface photos tagged as "whiteboard," but it will not read the handwriting back to you or tell you which of the eleven whiteboard shots contains that specific sentence. You are back to scrolling and squinting, which is the exact problem describing a photo was supposed to solve.

Pro Tip

If you rely on your camera roll as a memory bank for passwords, receipts, and notes, test any search tool with a real query like "the Wi-Fi password from the apartment" or "the whiteboard with the launch dates." A tool that only filters thumbnails will make you scan; a tool that reads the text and cites the source will hand you the answer.

What actually fixes it: one searchable memory across everything

The fix is to stop searching one gallery at a time and instead index everything you own in one place. That means running OCR on every image, screenshot, PDF, and scanned page so the text inside becomes searchable, transcribing voice notes so spoken words count too, and building a semantic index so a plain-English question can match the right item regardless of which app it started in. The goal is not a filtered grid. It is a direct answer that points to the exact source.

This is the specific job MemX does. Ask MemX for the Wi-Fi password and it returns the answer next to the exact screenshot it read. Ask for the whiteboard's next steps and it points at the right photo. You save photos, screenshots, PDFs, scanned documents, voice notes, and WhatsApp messages, and MemX runs OCR and transcription across all of it, indexes it together, then answers your questions in plain English with a citation to the source document. It works on Android, iOS, and WhatsApp, so it is not fenced to one brand of phone, one chip generation, or a short country list. On privacy, MemX is private by architecture: per-user data isolation, customer-managed encryption keys, encryption at rest, on-device capture, your content is never used to train AI models, and you can export everything anytime.

Insight

The pain in this post is finding one specific image across everything you own, not just inside one gallery. That is the gap MemX is built to close: OCR and search over your photos, screenshots, and documents together, with the source image returned next to the answer.

Frequently asked questions

Frequently Asked Questions
01How do I find a photo by describing it on my phone?

Open your gallery's search and type what is in the image, like "receipt" or "dog on beach." Full descriptive phrases work only on newer hardware: iPhone needs Apple Intelligence (iPhone 15 Pro or later), and Pixel's screenshot search needs a Pixel 9 or 10. Cross-app tools like MemX add this on any phone.

02Can I search my iPhone photos by what is in them?

Yes. Basic search for people, pets, objects, scenes, and text via Live Text works on recent iPhones. Full natural-language phrase search, such as "couple dancing at wedding," needs Apple Intelligence, which requires an iPhone 15 Pro, iPhone 16, or iPhone 17 running iOS 18.1 or later.

03Does Google Pixel let you search screenshots by description?

On supported models, yes. The Pixel Screenshots app allows natural-language search of screenshots, but only on Pixel 9 and Pixel 10 phones, only in English, German, or Japanese, and only in ten listed countries. It searches screenshots, not your full camera roll, PDFs, or documents.

04Why can't my phone find a screenshot I saved from a chat app?

Native photo search usually indexes only the Photos or gallery library. A screenshot saved inside a messaging app or downloaded elsewhere can sit outside that index, so it never appears in results. A tool that indexes all your images and files together, like MemX, can still find it.

05Is there a way to search photos and PDFs together?

Not with native phone galleries, which search images only and ignore documents. MemX runs OCR across your photos, screenshots, PDFs, and scanned pages and indexes them together, so one plain-English question can return the right item and cite the source, whether it is a photo or a document.

Read Next

Or try MemX to access 40+ AI models in one place — including Claude Sonnet 4.6 and GPT-5.4 — and get your questions answered today.

Was this article helpful?

Found this useful? Share it with someone who needs it.

Free · iOS, Android & WhatsApp

Stop losing what you save.
Let MemX remember it for you.

Every screenshot, photo, PDF and voice note — captured, encrypted, and instantly searchable. Ask in plain English, get the answer in seconds.

  • Reads text inside images and handwriting
  • Private and encrypted by default
  • Free to start, no credit card

Takes under a minute to set up. Your data stays yours.

Arpit Tripathi
Written by
Arpit TripathiLinkedIn

Founder of MemX. Ex-Google Staff Tech Lead Manager, ex-AWS Senior SDE (Elastic Block Store). Writes about practical AI on the MemX blog.

Keep reading

More guides for AI-powered students.