The short answer
Pick the clearest single-person photo you have, crop it to the face, and run it through Google Lens, then TinEye and Bing. You're looking for the same photo on a profile or page you don't recognise. If the person uses different photos on different apps, reverse image search will usually miss them, and you'll need a face search, which looks for the same face rather than the same picture.
Which photo to use (and how to prep it)
The photo you choose matters more than the engine you use. A good photo on a basic engine beats a bad photo on the best one.
- For finding a partner's profile: use the kind of photo they'd put on a dating app, such as a flattering, front-on, well-lit shot. Their profile pictures on Instagram or WhatsApp are good candidates.
- For checking a new match: use the photos from their dating profile itself. Screenshot each one, crop away the app interface, and search them one at a time.
- Crop to one person. Group shots confuse every engine.
- Use the original if you can. Every screenshot and re-save loses detail.
Check your photo here before you search. It runs in your browser and nothing is uploaded. Our other free tools work the same way.
Google Lens, TinEye, Bing, Yandex: what each is good at
| Method | Coverage | Exact copies | Similar faces | Ease |
|---|---|---|---|---|
Google Lens lens.google.com or the Google app | ||||
TinEye tineye.com | ||||
Bing Visual Search bing.com/visualsearch | ||||
Yandex Images yandex.com/images |
- Google Lens has the largest index and is the best first stop. It mixes exact matches with "visually similar" results, so read carefully. Google says it has been cautious about deploying face recognition, and Lens matches pictures rather than people.
- TinEye only finds copies of the same image, including edited ones. It can sort results by oldest, which is useful for spotting stolen photos. The date shown is when TinEye first found the image, not when it was posted, so read it as "at least this old".
- Bing returns pages that use the image and sometimes finds ones Google misses. It's worth a second run.
- Yandex was singled out by Bellingcat in 2019 as far better at matching faces than Google or Bing. Engines change, so don't count on that today. It's also run from Russia, so think about the privacy implications before uploading someone's photo.
Step by step
- 01
Prep
Crop to one face. Remove app overlays. Save as JPG or PNG.
- 02
Google Lens
Upload, then open 'Exact matches' and scan visual results for the same picture.
- 03
TinEye
Upload and sort by 'Oldest' to see the earliest copy TinEye has found.
- 04
Bing
Upload for a second opinion, especially if Google found little.
- 05
Record
Note every hit with its link and the date you found it.
Repeat for two or three different photos. Different pictures surface different results, and a single photo is easy to miss.
Doing it on a phone
Most dating-app photos are on your phone, so it helps to know the mobile shortcuts:
- iPhone: install the Google app and tap the camera icon in the search bar to use Lens on a screenshot. In Safari, you can also long-press some web images to search them.
- Android: Google Lens is usually built in. Open a screenshot in Google Photos and tap the Lens button, or long-press an image in Chrome and choose to search it with Google.
- TinEye and Bing work in a mobile browser. Upload from your photo library.
Before uploading a screenshot, crop it in your photo editor so only the photo remains, without the app's name, buttons or bio text. Overlays confuse matching.
Checking a new match before you meet
Reverse image search is also the quickest way to check that a new match is who they say they are. The US Federal Trade Commission recommends exactly this: reverse image search a suspected scammer's profile picture and see whether it's tied to another name. The FBI's internet crime centre notes that most romance scammers use someone else's photos, usually taken from social media.
- Screenshot each of their profile photos and crop them.
- Run each through Google Lens and TinEye. On TinEye, sort by oldest.
- Red flags: the photos appear under a different name, on a model or influencer account, on stock-photo sites, or on scam-warning pages.
- Reassuring signs: the photos only appear on their own social profiles, under the same name, with a consistent history.
Finding nothing is normal for many genuine people, and it doesn't clear a match on its own. The Illinois Attorney General warns that scammers increasingly use AI-generated images, which may not show up in a reverse image search at all because no earlier copy exists. If you're still unsure, a short video call before meeting is the most reliable check there is, and genuine matches rarely refuse one more than once.
A worked example: reading a TinEye result
Say a new match, "Sam, 31", has four photos. You crop the clearest one and run it through TinEye, sorted by oldest. Three results come back:
- An Instagram account under a different name, first found by TinEye four years ago, with hundreds of similar professional shots.
- A forum thread titled "fake profile using these photos" from last year.
- A dating aggregator page under the name Sam, first found last month.
The oldest copy belongs to someone else, a warning page already exists, and the "Sam" copy is the newest. That pattern points strongly to stolen photos. Unmatch, report the profile in the app, and don't send money or personal details. If the oldest copy had instead been on Sam's own long-running account under the same name, that would be reassuring.
Face search engines and how they differ
Face search engines detect the face in your photo, turn it into a numeric description of its features, and compare it against faces they've indexed from public web pages. Because they compare faces rather than pictures, they can find someone in a photo you've never seen, which is exactly what you need when a partner uses different pictures on a dating app.
They index different parts of the web, which changes what they can find. PimEyes says it doesn't search social media or other login-restricted sources, and its free searches only show that matches exist, with source links behind a paid plan. FaceCheck.ID, which is the engine CheaterFinder uses, also charges to search.
The trade-off is lookalikes. Every result comes with a similarity score, and high scores between relatives or strangers with similar features are common. FaceCheck.ID's own guide labels 90–100 a certain match, 83–89 confident, 70–82 uncertain and 50–69 weak, and warns that many unrelated people look alike. Treat anything below 70 as a probable lookalike, and even 90+ as needing a second detail. We explain scoring in depth on our search by photo page.
CheaterFinder face-matches the first photo you add, and combines it with name, location and username searches so face matches come with context. How it works shows each step, and we've set out what we do and don't do if you'd like to check us before paying.
Reading results: exact copy, edited copy, lookalike
Exact copy
Same picture, same crop. The photo has been posted elsewhere.
Tell: Identical background, clothes and expression.
Edited copy
Same picture, but cropped, flipped, filtered or resized.
Tell: Same moment, but different framing or colour.
Same person
Different photo, same face. Only face search finds these reliably.
Tell: Different setting, matching features, moles or tattoos.
Lookalike
Different person with a similar look, pose or style.
Tell: Close at a glance, but details like ears, teeth and hairline differ.
The tells are in small details. Lookalikes usually differ in ears, hairline, teeth, eyebrow shape, or moles and scars. The same person shows the same small details across different photos. If you're unsure, it isn't a match yet.
Why you sometimes get nothing
No results is very common, and it usually means one of these:
- The photo was never posted publicly. Photos taken for a dating app often exist only in that app.
- Social apps aren't fully indexed. Instagram, Facebook and TikTok limit what search engines can see, especially for private accounts.
- The photo was edited. Heavy filters, mirroring or tight crops can defeat exact-match engines.
- Image quality is too low. Small, blurry screenshots lose the detail engines need.
- The image was generated. An AI-made face has no earlier copies to find.
Next steps: try a different, higher-quality photo, try a face search instead, or combine the photo with a name, city and username so other searches can do the work. See every method, ranked.
Common mistakes
- Treating "visually similar" as a match. Lens shows people with a similar look. Those are suggestions, not results.
- Searching only one photo. Different photos surface different pages. Use at least two.
- Searching the whole screenshot. App interface and text drown out the face.
- Stopping at the first hit. The first hit may be a copy of a copy. TinEye's "oldest" sort helps find the original.
- Reading TinEye's date as the posting date. It's when TinEye found the image. The photo may be older.
- Acting on a face alone. Stolen photos mean a matching face can belong to someone else's fake profile.
Privacy and ethics of searching faces
Searching a face is more powerful than searching a name, and it deserves more care. Search people whose identity genuinely affects your relationship or your safety, such as a partner or a match you're about to meet. Don't search strangers out of curiosity, exes who've moved on, or anyone you intend to contact against their wishes.
Some places have specific rules on biometric data like faces. Illinois' Biometric Information Privacy Act applies to any "private entity", which it defines to include individuals. Texas and Washington's laws focus on collection for a commercial purpose. In the EU, the GDPR doesn't apply to purely personal or household activity. How these rules apply to one person running a search is not settled, so our guide to the legal lines covers the basics.
If you find your own photos on a profile
Sometimes the search turns up something unexpected: your photos on a profile you didn't make. It's more common than people think, because scammers harvest photos from public social accounts. If it happens:
- Screenshot the profile, including the name and any link, and note the date.
- Report it for impersonation inside the app if you have an account, or contact the app's support through its help centre if you don't.
- Ask friends who use the app to report it too. Several reports often get faster action.
- Make your social accounts private, or remove the photos that were copied, to reduce repeat use.
- If your face shows up in face search results you don't want, some engines let you opt out. PimEyes, for example, has an opt-out form that removes your face from its results, though not from the original websites.
It's worth checking your own photos once in a while, the same way you'd check anyone else's.
What a hit proves
Combine a photo hit with at least one other independent detail before you treat it as confirmed. Then read what to do after you find a profile. More on photos, signs and next steps is in our guides, and common questions about our own search are in the FAQ.
FAQ
Can reverse image search find Tinder profiles?+
Rarely directly, because Tinder profiles aren't usually indexed. It can find the same photos on other public sites, which can lead to the person or reveal that the photos are stolen.
Which reverse image search is best for dating profiles?+
Google Lens for coverage, TinEye for finding the earliest copy it has seen of an image, and Bing as a second opinion. For finding the same person in different photos, you need a face search engine.
Does the person know if I reverse image search their photo?+
No. Reverse image search engines don't notify anyone.
Why does reverse image search show people who look similar but aren't the same?+
Engines like Google Lens also return 'visually similar' images: same pose, colours or setting. Those aren't matches. Look only at results showing the same picture, or clearly the same person.
Can I reverse image search a screenshot?+
Yes, but crop it tightly to the photo first and remove any app interface. Screenshots lose quality, so use the original image if you can.
Sources
- Our approach to facial recognition. Google AI. ai.google/facial-recognition/. Accessed November 2026.
- Can I sort my results?. TinEye Help. help.tineye.com/article/246-can-i-sort-my-results. Accessed November 2026.
- What does TinEye “first found on” mean?. TinEye Help. help.tineye.com/article/248-what-does-tineye-first-found-on-mean. Accessed November 2026.
- Using Bing Visual Search. Microsoft Support. support.microsoft.com/en-us/bing/using-bing-visual-search. Accessed November 2026.
- Guide to Using Reverse Image Search for Investigations (2019). Bellingcat. bellingcat.com/resources/how-tos/2019/12/26/guide-to-using-reverse-image-search-for-investigations/. Accessed November 2026.
- What to Know About Romance Scams. US Federal Trade Commission. consumer.ftc.gov/articles/what-know-about-romance-scams. Accessed November 2026.
- Cyber Actors Use Online Dating Sites to Conduct Confidence/Romance Fraud (PSA, 2019). FBI Internet Crime Complaint Center. ic3.gov/PSA/2019/PSA190805. Accessed November 2026.
- Consumer alert: Valentine’s Day tips to prevent romance scams (13 February 2026). Illinois Attorney General. illinoisattorneygeneral.gov/news/story/consumer-alert-attorney-general-raoul-offers-valentines-day-tips-to-prevent-romance-scams-2-13-26. Accessed November 2026.
- FaceCheck FAQ. FaceCheck.ID. facecheck.id/en/Face-Search/FAQ. Accessed November 2026.
- PimEyes FAQ. PimEyes. pimeyes.com/en/faq. Accessed November 2026.
- How to remove your images from PimEyes search results. PimEyes. pimeyes.com/en/blog/how-to-remove-your-images-from-pimeyes-search-results. Accessed November 2026.
- Opt-Out Request Form. PimEyes. pimeyes.com/en/opt-out. Accessed November 2026.
- Biometric Information Privacy Act (740 ILCS 14). Illinois General Assembly. ilga.gov/legislation/ilcs/ilcs3.asp?ActID=3004&ChapterID=57. Accessed November 2026.
- Regulation (EU) 2016/679 (GDPR), Article 2 and Recital 18. EUR-Lex. eur-lex.europa.eu/legal-content/en/TXT/?uri=CELEX%3A32016R0679. Accessed November 2026.