Circle to Search in 2026: Multi-Object and Whole-Outfit Search, Explained

0

๐Ÿ”Ž Reverse Image Search

Image search techniques made simple: upload a photo or paste a URL and search 6 engines at once. 100% free, nothing stored.

๐Ÿ”’ Your image is never stored on our servers longer than 24 hours and is never shared.

Circle to Search changed in February 2026. Until then it searched one thing at a time: you circled a shoe, a landmark or a line of text, and Google searched that one thing. The February update, built on Gemini 3, lets you search several objects in one image at once, a whole outfit, a whole room, a whole school of fish, and adds a virtual try-on step for clothes. This guide covers only what is new: the features, which phones get them, and how to use them. The basic gesture and setup are covered in our Circle to Search guide.

Direct answer: The main circle to search new features are multi-object search, Find the Look whole-outfit search and in-search virtual try-on, announced by Google on 25 February 2026 and powered by Gemini 3. They launched on the Samsung Galaxy S26 series and Pixel 10 phones; Google’s help page now also lists the Pixel 10a and Pixel 11 family, with wider rollout promised.

Circle to Search new features: what changed in 2026?

Circle to Search is Google’s on-screen search for Android: long-press the Home button or navigation handle, then circle, highlight, scribble or tap anything on your screen to search it without leaving the app. It launched on 31 January 2024 on the Pixel 8 and Galaxy S24 series, and Google says it now runs on more than 580 million Android devices, with users making billions of queries a month.

The February 2026 update is the biggest change since launch, and it changes what the feature is for. Before it, Circle to Search answered one question about one object. After it, the feature can look at a whole scene. Google’s announcement, written by Harsh Kharbanda, Director of Product Management for Search, describes the update as powered by “Gemini 3’s agentic planning, reasoning and tool capabilities”. In plain terms, the model now decides for itself which parts of your screen matter, crops them, runs several searches at once, and cross-references the results into one answer.

Three things arrived with that update, and they are the whole subject of this article:

  • Multi-object search, circle several things in one image and search them together, instead of repeating the gesture for each one.
  • Find the Look, circle a whole outfit and get the pieces broken down into shoppable matches.
  • Try It On, try a found garment on a photo of yourself, or on a model, without leaving the search overlay.

Around the same update, the search bar moved from the top of the screen to the bottom, with an “Ask anything” field that makes the experience feel more like a chat. That small change matters: you can now circle something and type a follow-up question in the same place.

What is multi-object search, and how does it work?

Multi-object search is the ability to select several things in one image and ask about all of them in a single search. The old version searched the one thing inside your circle. The new version searches everything you circled, together, and explains how the pieces relate.

Google’s own example is a photo of colourful fish: circle all the fish and ask “what are all these fish, and how do they coexist?” Circle to Search identifies each species, Google names the Honeycomb Filefish and the Moon Jellyfish in its announcement, then explains the science behind their underwater community and links out to the web for more. That is the difference in one sentence: the old version named one fish, the new version answers a question about the whole scene.

Under the surface, Google says the feature uses a technique it calls visual query fan-out. Instead of running one search for one match, the model plans a multi-step search: it picks out the most important parts of the image, crops them, runs several searches in parallel, and cross-references what it finds before compiling a final response, including images from across the web, for each item you searched.

Three uses make the most sense in practice:

  • Outfits. Circle a whole look and get every piece at once, instead of searching the jacket, then the shoes, then the bag.
  • Rooms. Circle a living room and build a mood board of the lamp, the chair and the rug in one pass.
  • Scenes with several things you cannot name. A market stall, a reef, a shelf of tools, anywhere the question is “what is all of this?” rather than “what is this one thing?”

The results are only as good as the image. Google’s help page advises choosing images where clothing and objects are clearly visible, and avoiding outfits with multiple layers or partially covered elements. That advice applies to multi-object search generally: if two objects overlap heavily, the model has to guess where one ends and the other begins.

How does Find the Look work?

Find the Look is the shopping face of multi-object search. Circle an entire outfit and Circle to Search breaks it into its components, jacket, top, trousers, shoes, accessories, and finds similar items for each one, all in the same search. Google’s help page gives the exact steps:

  1. On your phone, circle the entire outfit you want to search.
  2. Tap Find the look. Google breaks down every piece of the outfit.
  3. In the search results, scroll down to find individual matches for every part of the outfit.

If the Find the Look button does not appear, Google’s own workaround is to type “find the look” into the search bar and tap enter. The button appears when the feature recognises an outfit in your selection; typing the phrase triggers the same breakdown manually.

A few practical points, all from Google’s current help page:

  • Circle the whole outfit, not one garment. Circling just the jacket gives you jacket results; circling the person gives you the look.
  • Pick a clear, front-on photo where each item is visible. Layered outfits and partly hidden items are the cases Google warns about.
  • Results are similar items for sale, not necessarily the exact garment. A match means “visually similar and shoppable”, and prices and stock come from the retailer, not from Google.
  • The feature needs the phone to support it. Which phones those are is the next-but-one section, and it is a shorter list than the list of phones with basic Circle to Search.

Find the Look is also where the February update’s shopping numbers come from: Google says shopping-related searches are among the top uses of Circle to Search, which is why the outfit flow got its own button instead of being left as a generic multi-object search.

What is Try It On inside Circle to Search?

Try It On is a virtual dressing room inside the search results. Once Find the Look has broken an outfit into pieces, you can tap a garment, then tap Try it on, and see how it looks on a photo of yourself or on a model, without opening a shopping app. Google’s help page gives these steps:

  1. On your phone, use Circle to Search to find a clothing item.
  2. In the search results, tap an eligible product or image.
  3. Tap Try it on. If you do not see the button, that item is not eligible to try on.
  4. Give it time to load to see how the clothing looks virtually. The first time, follow the prompt to upload a photo of yourself as a reference or choose a model.

Try It On has more conditions than any other part of the update, and they are worth knowing before you look for the button:

  • Age and account. You must be 18 or older, signed in to a Google account, and have Web & App Activity and personalised recommendations turned on. Google’s help page states all three plainly.
  • Countries. It is currently available in Australia, Canada, India, Japan, the UK and the US. If you are elsewhere, the button will not appear however new your phone is.
  • Eligible items only. Lingerie, swimming costumes and accessories are not supported, and products under “Sponsored” shopping ads never get a Try It On button. The button appears only where Google’s try-on feature already works for that product listing.
  • It shows a look, not a fit. The tool renders the garment onto a photo of you or a model. It shows style, colour and proportions on a body; it does not measure you, and it cannot tell you whether a size will fit. Treat it as a style check, not a fitting room. Our find an item by image guide covers the wider question of judging a listing from its photos.

Which phones get the new Circle to Search features?

This is the question most readers actually have, and it is the one Google’s own pages answer most precisely, so here is the feature-by-device matrix, built from Google’s Android help page and its February announcement, both checked on 12 November 2026.

Ranking asset: feature-by-device matrix

PhoneBasic Circle to SearchMulti-object search & Find the LookTry It OnCircle to Search in the Camera app
Pixel 6, 6 Pro, 7, 7 Pro, 8, 8 Pro, 9 seriesYesNo, not listed by GoogleNoNo
Pixel 10, 10 Pro, 10 Pro XL, 10 Pro FoldYesYes, launch device, Feb 2026YesNo
Pixel 10aYesYes, listed by Google, Nov 2026YesNo
Pixel 11, 11 Pro, 11 Pro XL, 11 Pro FoldYesYes, listed by Google, Nov 2026YesYes, only these models
Samsung Galaxy S26, S26+, S26 UltraYesYes, launch device, Feb 2026YesNo
Other Galaxy phones & tablets (Android 14, One UI 6.1+)YesNot listed yet, Google says “more devices soon”Not listed yetNo
Other Android phones with Circle to Search (Xiaomi, Honor, Motorola and others)YesOnly on devices that support AI Mode and AI OverviewsOnly on devices that support AI Mode and AI OverviewsNo
iPhoneNo, Android onlyNoNoNo

Three things in that table deserve a closer look. First, the gate has widened since launch: in February the new features were on the Galaxy S26 series and Pixel 10 only, and by November Google’s help page also lists the Pixel 10a and the whole Pixel 11 family. Second, the wording that matters for every other phone is Google’s phrase “devices that support AI Mode and AI Overviews”, that is the test Google now publishes, not a model list. Third, Circle to Search inside the Camera app is the narrowest feature of all: Pixel 11 series only, at the time of writing.

Two conditions apply on top of the model list. On a Pixel, the feature is switched on in Settings > System > Circle to Search; on a Galaxy phone, the switch is in Settings > Display > Navigation bar > Circle to Search, and Samsung says it is on by default on supported devices. And everywhere, the feature needs an internet connection and does not work in apps that block screen capture, such as banking apps.

What else changed in Circle to Search in 2026?

Multi-object search gets the headlines, but three smaller changes arrived with it and are easy to miss.

The chat-style bar. The search field moved from the top of the screen to the bottom, labelled “Ask anything”. Circle something, then type a question about it in the same overlay, the same gesture, but now a conversation. AI Overviews can appear from these text-plus-image queries, so a circled photo plus a typed question can return a written answer, not just image matches.

Translate as you scroll. Long-press, tap Translate, and the page or app translates as you scroll through it. Google’s help page notes it works for most web pages, messages and apps, but not for videos and GIFs.

A photo transparency check. Circle to Search can now check whether a photo is an original or has been modified in some way, using automatic detection. Google links this to its “get more info about an image” flow. It is a detection aid, not a verdict, a clean result does not prove a photo is untouched, and a flagged result does not say how it was changed. If that question matters to you, our how to reverse image search guide and our Google reverse image search guide show how to check a photo’s history properly.

What should you do when Find the Look does not appear?

This is the most common complaint about the update, and most of the time the answer is the device list, not a bug. Work through these in order:

  1. Check your model against the matrix above. If you have a Pixel 9 or an older Galaxy, the feature has not reached your phone yet, Google’s February wording was “coming to more Android devices soon”, and nine months later the published list is still the Pixel 10 and 11 families plus the Galaxy S26 series and AI Mode/AI Overviews devices.
  2. Update the Google app and your system software. Circle to Search rides inside the Google app and Android System Intelligence, and the new features arrive with those updates, not with a separate download.
  3. Check the country and account rules for Try It On. If Find the Look works but Try It On never appears, check that you are in Australia, Canada, India, Japan, the UK or the US, signed in, 18 or older, and that Web & App Activity and personalised recommendations are on.
  4. Try the typed fallback. Circle the outfit, then type “find the look” in the search bar and tap enter. If the button appears after that, the feature is on your phone and simply did not recognise the outfit from the gesture alone.
  5. Pick a clearer photo. Google’s own tips are to choose images where clothing is clearly visible and to avoid layered or partly covered outfits. A busy street photo with overlapping people is the hardest case; a single person, front-on, against a plain background is the easiest.
  6. Accept the limits. Try It On never appears on Sponsored products, and it does not support lingerie, swimwear or accessories. If the button is missing on those items, nothing is broken.

How does Circle to Search compare with Google Lens now?

Circle to Search is built on the same visual-search engine as Google Lens, but the two are used differently. Lens is an app and a camera: you point it at something, or upload a photo to it. Circle to Search is a gesture on whatever is already on your screen: you never leave the app you are in. The February update narrows the gap in one direction, Circle to Search can now do multi-object searches that used to need a Lens upload and a series of separate searches, but it does not replace Lens. Lens still works on any phone with the Google app, including iPhone, and it still handles uploads of photos from your camera roll. Circle to Search still only exists on select Android phones.

For shopping, the practical rule is simple: if the item is on your screen, circle it; if it is a photo you already saved, upload it to Lens. And if you want the same trick on a video, our Google Lens in YouTube Shorts guide covers the in-player version of the same idea.

Worked example: I ran a real outfit photo through the same search engines

One honest note first: this run had no Pixel 10 or Galaxy S26 to hand, so the on-device Find the Look taps were not run on a phone here. What was run, genuinely and on 12 November 2026, is the closest test available from a desk: the same outfit photo put through the web visual-search engines that power the feature, to see what the current single-object search does with a whole outfit, and what the new multi-object step would have to add.

The test photo is a Wikimedia Commons image (CC BY-SA, by PattayaPatrol, 8 July 2025): a young woman in a fitted bright pink dress standing beside a night-market food stall, with a vendor in a red shirt behind a cart of bottles and cooking equipment. It is exactly the kind of busy, multi-object scene the February update is built for, dress, earrings, stall, bottles, and the kind of photo where the old version has to choose one thing.

Google Lens by URL. Getting in took several attempts: Lens repeatedly returned Google’s “unusual traffic” page and a reCAPTCHA, and one attempt produced no checkbox at all. A later attempt, pasting the image link through Lens’s “Search by image” flow and solving a “select all images with cars” grid, finally reached the results page. What Lens returned proves the point of the update: it auto-selected only the woman in pink, a selection box on the right side of the photo, not the vendor or the stall, and its AI Overview said: “The selected region shows a woman wearing a bright pink halter-neck bodycon mini dress.” It then asked whether I wanted to buy a similar dress or wanted details about the scene. Under “Visual matches” it showed priced look-alikes of the same dress: Lulus, Jovani at $550, L*Space at $123, eBay at $89, Amanda Uprichard at $238 and Lucy in the Sky at $15, plus Princess Polly, REVOLVE, Fashion Nova and Hello Molly. Its “Exact matches” tab said “No exact matches found”, Lens did not find the Wikimedia original either.

Yandex Images by URL. Yandex loaded immediately, with no block and no sign-in. It tagged the photo only in Russian, “girls from Thailand”, “people in Thailand”, “Thailand population”, “street food” and “market in Thailand”, with no mention of the pink dress, the woman or any product. Its similar images were thematic rather than identical: Thailand and Pattaya night-market scenes, and prominently a different woman entirely, the Laos omelette lady from YouTube street-food videos. No shopping results appeared, and no Wikimedia Commons source for the photo showed in the first results.

The contrast is the lesson. Given the same photo, Lens read it as a dress to shop for and returned priced look-alikes; Yandex read it as a Thailand street-food scene and returned thematically similar night-market photos. Neither engine identified the exact image or its source. And both did the old job: one object, one interpretation. The new Circle to Search step, circling the whole outfit and getting every piece broken down at once, is precisely what neither web engine did here, which is why the feature matters and why it is worth checking whether your phone has it before assuming a Lens upload will do the same thing.

What is worth trying first?

If your phone is on the list, start with an outfit, not a single item, that is the one search the old version could not do at all. Circle the whole look, tap Find the Look, scroll the per-piece matches, and try one garment on. Then try the same photo as a multi-object scene question, such as asking what the items in a room are. If your phone is not on the list, there is nothing to install and no setting to force: the February update is still rolling out, and Google’s own list is the only honest way to check whether it has reached you.

Which phones get the new Circle to Search features?

Multi-object search, Find the Look and Try It On launched on the Samsung Galaxy S26 series and Pixel 10 phones in February 2026. Google's help page now also lists the Pixel 10a and the Pixel 11 family, plus any device that supports AI Mode and AI Overviews. The basic feature is wider: Pixel 6 and later.

Can I use Find the Look on an older Pixel or Galaxy?

Not yet, according to Google's published list. Find the Look needs a Pixel 10, Pixel 10a or Pixel 11 phone, a Galaxy S26, or another device that supports AI Mode and AI Overviews. A Pixel 9 or older Galaxy has basic Circle to Search but no Find the Look button, and Google has not published a date for wider rollout.

Is Try It On available in the UK?

Yes. The UK is one of the six countries where Try It On currently works, Australia, Canada, India, Japan, the UK and the US. You must also be 18 or older, signed in to a Google account, and have Web and App Activity and personalised recommendations turned on. The button still appears only on eligible products.

Does Circle to Search work on iPhone?

No. Circle to Search is an Android-only feature, built into the Google app and Android System Intelligence on select Android phones. iPhone owners cannot install it or switch it on. The closest equivalent on iPhone is the Google app's Lens button, which searches a photo you point the camera at or upload, or Lens inside YouTube Shorts.

Ali Butt is the lead writer and SEO specialist at Image Search Techniques, where he covers reverse image search, visual search, and online image verification. With over four years of experience in SEO and content, he writes practical, hands-on guides that help readers trace a photo to its source, identify people, products, plants and brands from a picture, protect their own images online, and spot catfish, fakes and AI-generated images across Google, Yandex, Bing, TinEye and more.

Leave A Reply