# FaceSearch > FaceSearch is a reverse face search engine that finds where a person's face appears online. Upload one photo and it scans social, dating, and public web sources for matches. ## Key facts - Category: Reverse face search engine - Input: one photo of a face. There is no text-query search. - Credits per search: 3 - Free credits on signup: 3, no credit card required - Uploaded images are deleted within 24 hours of search completion - Interface languages: 10 (en, zh, es, fr, de, ja, ko, pt, ru, ar) - Support: support@facesearch.net ## Topics covered - Reverse face search - Facial recognition search - Reverse image search - Catfish and romance scam detection - Online identity verification - Social media profile discovery - Image privacy and unauthorized photo use ## Core - [FaceSearch - Free Reverse Face Search | Find Anyone by Photo](https://facesearch.net): FaceSearch is a free reverse face search engine. Upload one photo to find matching faces across social, dating and web sources, verify identities and spot catfish. - [About FaceSearch | What Reverse Face Search Is and How It Works](https://facesearch.net/about): FaceSearch is a reverse face search engine. Learn what it does, how face matching works, what it deliberately does not do, how uploaded photos are handled, and who to contact. ## How reverse face search works - [What Is Reverse Face Search? How It Works and When to Use It](https://facesearch.net/reverse-face-search): Reverse face search finds where a person's face appears online by comparing facial geometry, not filenames. How it works, how it differs from image search, and its limits. - [Face Search vs Reverse Image Search: What Is the Difference?](https://facesearch.net/face-search-vs-reverse-image-search): Reverse image search finds copies of your file. Face search finds the person in it. What each one matches on, when to use which, and why results differ so much. - [Is Reverse Face Search Legal? Rules by Region, Explained](https://facesearch.net/is-face-search-legal): Whether reverse face search is legal depends on where you are and what you do with the result. How GDPR, BIPA and US state law treat face data, explained. - [How to Spot a Catfish: 9 Signs and How to Verify a Photo](https://facesearch.net/how-to-spot-a-catfish): Catfish profiles follow patterns. The warning signs in a profile and in conversation, how to verify a photo with reverse face search, and what to do once you know. - [Face Search Returned No Results: What It Means and What to Try](https://facesearch.net/face-search-no-results): An empty face search result is information, not a failure. What no results actually tells you, the photo problems that cause it, and what to change before searching again. - [Face Search Glossary: The Terms Explained Plainly](https://facesearch.net/face-search-glossary): Embedding, confidence score, false positive, biometric identifier, liveness. The vocabulary of reverse face search, defined in plain language with why each one matters. - [How Facial Recognition Works: Embeddings, Thresholds, Limits](https://facesearch.net/how-facial-recognition-works): Facial recognition turns a face into a list of numbers and compares distances. What a face embedding holds, what a match threshold decides, and what the technology cannot do. - [How to Remove Your Face From Face Search Results](https://facesearch.net/remove-face-from-search): FaceSearch keeps no face index to remove you from. Here is how removal actually works: the source sites, the face-search indexes, and your legal right to erasure. - [How Accurate Is Reverse Face Search?](https://facesearch.net/face-search-accuracy): What a face match accuracy figure means, what a confidence score measures, and why an accurate matcher still returns the wrong person at web scale. - [How to Find Someone Using a Photo](https://facesearch.net/how-to-find-someone-by-photo): The steps to find someone from a single photo: which image to use, what a face search can and cannot return, and how to check a match before acting on it. - [Reverse Video Search: Find Someone From a Video](https://facesearch.net/reverse-video-search): No engine searches a video file for a person. The working method is to take one clear frame and run a reverse face search on it. Here is how to pick that frame. - [How to Verify Someone You Met Online: A Step-by-Step Check](https://facesearch.net/verify-someone-you-met-online): A reverse face search is one check, not proof of identity. The steps that actually verify someone you met online, what a face match proves, and what it cannot. - [Reverse Face Search on a Phone: iPhone and Android](https://facesearch.net/face-search-on-phone): You can run a reverse face search entirely from a phone, from the camera roll or a screenshot. The steps on iPhone and Android, and why a phone search comes back empty. - [How Much Does a Reverse Face Search Cost?](https://facesearch.net/pricing): What a face search costs, what a credit buys, whether a genuinely free reverse face search exists, and how to choose between a monthly plan and a credit pack. ## Platform face search - [TikTok Face Search - Find TikTok Users by Photo | FaceSearch](https://facesearch.net/tiktok-user-finder): Find matching TikTok profiles from any photo. FaceSearch identifies creators, verifies influencers, and detects catfish accounts across platforms. - [Instagram Face Search - Find Users by Photo | FaceSearch](https://facesearch.net/instagram-user-search): Find matching public Instagram profiles from any photo. FaceSearch's Instagram face search identifies users, verifies influencers, and protects your brand. - [Facebook Face Search - Find Profiles by Photo | FaceSearch](https://facesearch.net/facebook-user-search): Uncover matching Facebook profiles from a single photo. FaceSearch scans public posts, Groups, and linked platforms to verify identities and expose fakes. - [Twitter/X Face Search - Find Profiles by Photo | FaceSearch](https://facesearch.net/twitter-user-search): Identify Twitter/X accounts from a photo. FaceSearch reads avatars and Spaces covers to attribute viral clips and track impersonators in seconds. - [LinkedIn Face Search - Find Profiles by Photo | FaceSearch](https://facesearch.net/linkedin-user-search): Verify a LinkedIn presence from any photo. FaceSearch confirms candidate and executive identities, flags duplicate accounts, and speeds compliance checks. - [Snapchat Face Search - Find Users by Photo | FaceSearch](https://facesearch.net/snapchat-user-search): Identify Snapchat users from a selfie or screenshot. FaceSearch matches faces across platforms to protect teens and expose copycat accounts. - [YouTube Face Search - Find Channels by Photo | FaceSearch](https://facesearch.net/youtube-face-search): Discover YouTube creators by face. FaceSearch scans thumbnails, channel art, and video frames to match any photo to the right YouTube channels. - [Social Media Finder - Find Anyone by Photo | FaceSearch](https://facesearch.net/social-finder): Find a person across all social media by photo. FaceSearch matches faces on Facebook, Instagram, TikTok, LinkedIn, X, and 100+ platforms instantly. - [FaceSearch - Reverse Face Search Engine](https://facesearch.net/facesearch-ai): Upload a photo and FaceSearch scans 100+ social, dating, and web sources with neural facial recognition to link identities and flag risk in seconds. - [Cheater Finder by Photo - Catch a Cheater | FaceSearch](https://facesearch.net/cheater-finder): Suspect infidelity? FaceSearch's cheater finder scans dating sites, social media, and the web by photo to uncover hidden profiles and reveal the truth. ## Policies and support - [Contact Us | FaceSearch](https://facesearch.net/contact): Reach the FaceSearch team for support, billing, partnership, or privacy questions. We're here to help with any request about reverse face search. - [Support Policy | FaceSearch](https://facesearch.net/support): How FaceSearch handles support requests, billing questions, technical issues, and account help: response times, contact channels, and what to expect. - [Refund Policy | FaceSearch](https://facesearch.net/refund-policy): FaceSearch refund rules for credits, subscriptions, completed searches, technical failures, and payment disputes: what qualifies and how to request one. - [Privacy Policy | FaceSearch](https://facesearch.net/privacy): How FaceSearch collects, uses, stores, and protects your data, including photo handling, biometric processing, data retention, and your privacy rights. - [Terms of Service | FaceSearch](https://facesearch.net/terms): The terms governing your use of FaceSearch: acceptable use, account rules, billing and credits, liability, and the responsible-use policy for face search. ## Optional - [Full text](https://facesearch.net/llms-full.txt): every page description, key facts, and the full FAQ in one file. - [Sitemap](https://facesearch.net/sitemap.xml): all 31 pages across 10 locales, with hreflang. - Other languages: [zh](https://facesearch.net/zh), [es](https://facesearch.net/es), [fr](https://facesearch.net/fr), [de](https://facesearch.net/de), [ja](https://facesearch.net/ja), [ko](https://facesearch.net/ko), [pt](https://facesearch.net/pt), [ru](https://facesearch.net/ru), [ar](https://facesearch.net/ar) ## Verified profiles - https://apps.apple.com/us/app/facesearch-face-image-search/id6795353307 - https://play.google.com/store/apps/details?id=net.facesearch.app - https://github.com/face-search/facesearch-skill - https://theresanaiforthat.com/ai/facesearch/ ## Frequently asked questions ### How does FaceSearch work? FaceSearch is a specialized face search engine powered by advanced facial recognition technology. Simply upload a photo, and our system performs a deep scan across social media and websites to find matching faces. Our AI analyzes facial features and searches millions of sources to deliver accurate results. ### What happens to my uploaded images? Your privacy is our top priority. Uploaded images are processed securely using encrypted connections and are automatically deleted within 24 hours after search completion. We never store face data permanently or share it with third parties. Your data remains yours alone. ### How are credits consumed? Each face search requires 3 credits. Credits are reserved when you submit a search and consumed upon successful completion. If a search fails due to a technical error on our end, your credits are automatically refunded. Please note that searches returning 'no results' still consume credits as the full search process was completed. ### What payment methods are accepted? We accept all major credit and debit cards (Visa, Mastercard), digital wallets (Apple Pay, Google Pay), PayPal, and regional payment methods like Alipay and WeChat Pay depending on your location. We also support over 300 cryptocurrencies for secure, anonymous payments. ### Can I get a refund? Credits are non-refundable once purchased. However, if you experience technical issues that prevent searches from completing, our dedicated support team will help restore your credits. PRO subscriptions can be cancelled at any time, and you'll retain access until the end of your billing period. ### Is this legal to use? FaceSearch is a powerful tool that must be used responsibly and in compliance with all applicable local laws and regulations. Users must confirm they have the legal right to search each image. We strictly prohibit using our service for stalking, harassment, or any illegal activities. ### How accurate are search results? Our AI-powered system achieves approximately 95% match accuracy under optimal conditions. However, accuracy depends on various factors including image quality, camera angle, lighting conditions, and how frequently the face appears online. We always recommend manually verifying important results for complete confidence. ### Can I search for anyone? You can search for faces that appear publicly online. However, you must have the legal right to search for the image (for example, it's your own photo, you have explicit permission, or it's for legitimate business purposes like employment verification). Always respect privacy laws and regulations in your jurisdiction. ### What types of images work best? For optimal results, upload clear, well-lit photos with the face clearly visible and looking directly at the camera. Avoid blurry images, extreme angles, or photos where the face is partially covered by accessories or shadows. Recent selfies and professional photos typically deliver the best results. ### How long does a search take? Most searches complete within 1-3 minutes. The exact processing time depends on the complexity of the search, image quality, and the number of sources being scanned. You'll receive a notification as soon as your comprehensive results are ready for review. ### What does the PRO plan include? The PRO plan unlocks unlimited face searches, complete access to search history, professional PDF export functionality, priority processing for faster results, and unlimited result downloads. It's designed for professionals and power users who require comprehensive search capabilities for frequent use. ### Can I download the search results? Absolutely! PRO users can export complete search results as professional, branded PDF reports containing all matches, confidence scores, and source links. Free users can view results online and manually save individual images. PDF exports are perfect for documentation, legal purposes, or sharing. ### What if no results are found? If no matches are found, this typically means the individual has no public online presence, or the image quality was insufficient for accurate matching. We recommend trying different photos with better angles, lighting, or clarity. Credits are still consumed as the full search process was completed. ## What Is Reverse Face Search? > Reverse face search finds other places online where the same person's face appears. You give it one photo, it converts the face into a numeric descriptor, and it compares that descriptor against faces on publicly reachable pages. It matches on facial geometry alone, never on the filename, the caption, or any text near the image. In short: - The input is a face, not a keyword. There is no way to search by name. - It compares facial geometry, so it finds the same person in a different photo, at a different age, in different lighting. - Reverse image search finds copies of your file. Face search finds the person. - A match is a lead worth checking, never proof of identity. - Only publicly reachable pages are searchable. Nothing behind a login is. ### How does reverse face search work? A face detector first locates the face in your photo and normalises it for angle and scale. A recognition model then turns that face into an embedding: a list of numbers describing the geometric relationships between features, such as the relative distance between the eyes and the shape of the jawline. Those relationships stay roughly stable when lighting, expression, hairstyle, and age change, which is why the same person can be matched across photos taken years apart. The embedding is compared against an index built from publicly reachable pages, and the closest matches come back ranked by a similarity score. - Detect: find the face and correct for angle and scale. - Encode: convert the face into a numeric embedding. - Compare: measure distance between that embedding and indexed ones. - Rank: return the nearest matches with a similarity score. ### How is it different from reverse image search? Reverse image search looks for your image. Face search looks for your face. Upload a photo to a reverse image engine and it finds copies and near-copies of that exact file: the same picture reposted, cropped, or resized. Give the same photo to a face search engine and it finds other photos of the same person, which may look nothing like the original. That difference decides which tool answers your question, and the two are not interchangeable. ### What is reverse face search used for? The common thread is verifying that a person online is who they claim to be, or finding out where your own face has ended up. It answers questions a name search cannot, because names are easy to change and faces are not. - Checking whether a dating or social profile belongs to a real person. - Finding out whether your photos are being used on accounts you do not control. - Confirming that a public-facing professional profile is genuine. - Attributing a photo or video to the person actually in it. ### How accurate is it? Accuracy depends far more on the photo you supply than on the engine. A sharp, front-facing, well-lit image of one unobscured face is the best case. Group shots, heavy filters, sunglasses, masks, extreme angles, and low resolution all degrade the embedding and therefore the match. Scores below the confidence threshold indicate a possible match, and treating a possible match as a confirmed identity is the single most common way people misuse these tools. ### What can reverse face search not do? Reverse face search cannot confirm identity, and no engine of this kind can. It reads only what is publicly reachable, so private accounts, closed groups, and anything behind a login are invisible to it. It does not search government, medical, financial, or criminal records, it cannot search by name or phone number, and it does not tell the person that they were searched for. ### Is reverse face search legal? Whether reverse face search is legal depends on where you are and what you do with the result. Facial recognition is regulated very differently across jurisdictions: some treat face data as sensitive biometric data requiring consent, others regulate the downstream use rather than the search itself. Verifying someone you are already talking to, or looking for misuse of your own photographs, sits differently from building a profile on a stranger. Source: What Is Reverse Face Search? How It Works and When to Use It ## Face Search vs Reverse Image Search > Reverse image search matches the picture. Face search matches the person. Give both tools the same photo and reverse image search returns copies of that file, while face search returns other photos of the same face, which may look completely different. Choosing the wrong one is the usual reason people conclude, incorrectly, that there is nothing to find. In short: - Reverse image search compares pixels and visual fingerprints of the file. - Face search compares facial geometry and ignores the rest of the picture. - A cropped or recoloured repost still matches in image search. - A completely different photo of the same person only matches in face search. - If the photo has never been posted anywhere else, image search finds nothing while face search still can. ### What does reverse image search match on? Reverse image search builds a fingerprint of the file itself: colour distribution, edges, and overall composition. It is designed to answer where else this exact picture appears, and it is good at surviving resizing, mild cropping, and recompression. What it cannot do is recognise that two different photographs show the same person, because visually those two files have almost nothing in common. ### What does face search match on? Face search throws away almost everything reverse image search relies on. It locates the face, normalises it, and reduces it to a numeric description of facial geometry. Background, clothing, colour grading, and resolution are discarded. That is why it can connect a studio headshot to a candid holiday snapshot, and equally why it will not help you find where a landscape photo was reposted. ### When should you use which? Ask what your actual question is. If the question is about a picture, use reverse image search. If the question is about a person, use face search. - Where was my photograph reposted? Reverse image search. - Is this dating profile using someone else's photos? Face search. - Who took this image originally? Reverse image search. - Does this person have other accounts under a different name? Face search. - Is this product photo stolen from my shop? Reverse image search. ### Can you use both together? Yes, and for most investigations you should. Run reverse image search first to find where that exact file has been reposted, which often surfaces the original source and its date. Then run face search to find accounts using different photographs of the same person. The two produce different evidence, and a claim supported by both is much stronger than one supported by either alone. ### Why do the two return such different result counts? Image search and face search are looking for different things, so the counts are not comparable. A widely reposted stock photo can return thousands of image-search hits and almost no face-search hits, because the same file is everywhere but the person is not. A private individual's own photo can return zero image-search hits and several face-search hits, because that exact file was never reposted while the person has other public photos. Neither number measures how much is out there about the person. Source: Face Search vs Reverse Image Search: What Is the Difference? ## Is Reverse Face Search Legal? > In most places, running a reverse face search is not illegal in itself, but what you do with the result can be. Face data is treated as sensitive biometric data under GDPR and several US state laws, and the deciding factors are usually your purpose, your jurisdiction, and whether the person consented. This page explains the shape of the rules. It is not legal advice, and if a decision carries real consequences you should take advice for your own jurisdiction. In short: - The search itself and the use of the result are regulated separately. - GDPR treats facial recognition data as a special category needing an explicit legal basis. - Illinois BIPA is the strictest US regime and carries a private right of action. - Verifying someone you are already dealing with is treated very differently from profiling a stranger. - Stalking, harassment, and unmasking someone against their will are unlawful almost everywhere. ### What exactly do these laws regulate? Most face-related privacy law regulates the processing of biometric identifiers rather than the act of looking at a photograph. A face embedding, the numeric descriptor a face search engine computes, is generally treated as a biometric identifier because it can single out one individual. That framing matters: a rule may be triggered by the computing and storing of that descriptor, independently of whether you ever act on the result. ### How does GDPR treat face search in Europe? Under the GDPR, biometric data processed for the purpose of uniquely identifying a person is a special category of personal data under Article 9, and processing it is prohibited unless a specific exception applies, most commonly the individual's explicit consent. The UK GDPR takes materially the same approach. There is a narrow household exemption for purely personal activity, but regulators have read it narrowly, and it is generally understood not to cover publishing or acting on results about other people. ### What about the United States? There is no single federal biometric privacy statute, so the answer is state by state, and the differences are substantial. - Jurisdiction Illinois. Statute: BIPA (740 ILCS 14); Consent required: Written and informed, before collection; Enforced by: Individuals, directly. - Jurisdiction Texas. Statute: CUBI (Bus. & Com. Code 503); Consent required: Yes; Enforced by: State attorney general. - Jurisdiction Washington. Statute: RCW 19.375; Consent required: Yes; Enforced by: State attorney general. - Jurisdiction California. Statute: CCPA / CPRA; Consent required: Treated as sensitive personal information; Enforced by: State agency and attorney general. ### What separates lawful use from unlawful use? Purpose is usually what decides it. Checking whether a person you are already dealing with is who they say they are, or finding where your own photographs have been reposted, are the uses regulators and platform terms treat most favourably. Building a dossier on a stranger, tracking someone's movements, revealing an anonymous person's identity, or feeding results into an employment, housing, credit, or insurance decision are where legal exposure concentrates, and some of those uses are separately regulated whatever the data source. ### What does FaceSearch require of you? Our terms require that you have the legal right to search each image you upload and that you comply with the law where you are. Using the service to stalk, harass, intimidate, or unmask a person against their will is prohibited and is grounds for termination. We do not permit use for automated decisions about employment, housing, credit, or insurance. Uploaded photos are deleted within 24 hours of the search completing, and face descriptors are not retained after deletion. Source: Is Reverse Face Search Legal? Rules by Region, Explained ## How to Spot a Catfish > A catfish is someone using another person's photos and a false identity to build a relationship online. The reliable tells are structural rather than emotional: a thin photo history, a refusal to appear live on camera, a story that explains every absence, and money entering the conversation. The fastest single check is running one of their photos through reverse face search to see whether that face belongs to someone else. The short version: - Refusing a live video call, repeatedly and with good excuses, is the strongest single signal. - Stock-quality photos with no ordinary, badly lit ones is a pattern worth checking. - Any request for money, gift cards, or crypto ends the question. Stop there. - A reverse face search on their photo takes a minute and settles most cases. - No match does not prove they are real. It only means those photos are not widely posted. ### What is catfishing? Catfishing is building a relationship online under an invented identity, usually with photographs taken from a real person who has nothing to do with it. Motive varies: some are after money, some are hiding an existing relationship, some are lonely and afraid the truth will end the conversation. The motive matters for how it ends, but not for how you detect it. The mechanics are the same in every case, because the person has to keep explaining why they cannot be seen or verified. ### What are the warning signs in a profile? Look at the shape of the photo set rather than at any single picture. Real accounts accumulate mess: bad lighting, other people, unflattering angles, years of drift in haircut and weight. - Very few photos, all clearly posed or professionally shot. - No photos with friends or family, or the same crop of a group shot repeatedly. - The account was created recently but claims a long history. - Follower and following counts that do not match the claimed life. - Details that shift between the profile and the conversation, such as age, city, or job. ### What are the warning signs in conversation? The pattern to watch for is escalating intimacy paired with escalating unavailability. Someone who is genuinely interested becomes easier to reach over time, not harder. - Declaring strong feelings unusually early. - A live video call is always about to happen and never does. - A profession that conveniently explains the absence, often overseas, offshore, or deployed. - A sudden crisis that only money can solve. - Pressure to move off the platform to private messaging immediately. ### How do you verify their photo? Run one of their clearest photos through reverse face search. If the same face appears on accounts under a different name, or on a stock photography site, or attached to a public figure, you have your answer in a minute. Use a photo where the face is large, front-facing, and unobstructed, because the check is only as good as the image you give it. - Pick the sharpest photo showing one unobscured face. - Run a reverse face search on it. - Check whether matches use a different name or a different life story. - Also run a reverse image search, which catches stock photos and reposts. ### What does it mean if the search finds nothing? Less than people assume. No result means those particular photographs are not widely posted on publicly reachable pages. That is consistent with an honest person who keeps a small online footprint, and it is equally consistent with a catfish who took photos from a private account. Treat an empty result as one piece of evidence that failed to resolve, not as a clearance, and keep asking for a live video call. ### What should you do if you are being catfished? Stop sending money immediately and do not send more in the hope of recovering what you have already sent, which is the most common way losses compound. Preserve the evidence before you act: screenshot the profile, the photos, and the conversation, because accounts disappear quickly once confronted. Report the account to the platform, and if money changed hands, report it to your bank and to your country's fraud reporting body. If intimate images are involved, that is a criminal matter in most jurisdictions and worth reporting to the police rather than handling alone. Source: How to Spot a Catfish: 9 Signs and How to Verify a Photo ## Your Face Search Returned No Results > An empty result usually means one of two things: the photo was too hard to read, or that face is genuinely not on publicly reachable pages. Those have different fixes, and the photo is worth ruling out first because it is the half you control. Change the image before you conclude anything about the person. The short version: - No results is a real answer, not an error, and the credits were spent doing the work. - Photo quality causes most empty results. Rule it out before drawing conclusions. - One sharp, front-facing, well-lit face beats a group photo every time. - Private accounts are invisible to any face search, so absence proves very little. - Trying a genuinely different photo is worth more than repeating the same one. ### What does no results actually mean? An empty result means no face on the indexed, publicly reachable pages scored above the confidence threshold against the face you supplied. It does not mean the person has no online presence, that they are not real, or that their photos are not out there somewhere. It is a statement about what was findable from that one image, which is a narrower claim than most people read into it. ### Did the search fail, or did it finish with no matches? A search that finishes and returns nothing is not a search that failed, and FaceSearch treats the two differently. A completed job means every reachable source was compared against your photo and no face scored above the confidence threshold: the search was working, no matches were found, and the credits stay spent because the work was actually done. A failed job is the other case — it stops with an error, the reserved credits are returned, and it can be retried. If your history shows a finished search with an empty result list, nothing is broken, and the rest of this page is about why the list is empty. ### Which photo problems cause empty results? Almost everything that makes a photo hard for a person to identify also makes it hard for a model. The face has to be large enough in the frame, lit well enough to show its structure, and turned close enough to the camera. - The face occupies a small part of the frame, so there is little detail to encode. - Sunglasses, a mask, a hand, or hair covering part of the face. - Strong beauty filters, which change the geometry the match depends on. - A steep angle, a profile shot, or the head tilted far back. - Harsh backlighting or deep shadow across one side of the face. - Heavy compression, screenshots of screenshots, or a very old low-resolution image. ### How do you pick the best photo? Choose the image where the face is largest and clearest, even if it is less flattering. A slightly awkward passport-style photo outperforms a beautiful one taken at a distance. Changing the image is the only reliable way to improve a face search, because it changes what the model has to work with. - One face, looking near the camera. - The face filling a meaningful part of the frame. - Even lighting on both sides of the face. - No sunglasses, no heavy filter, no obstruction. - The original file where possible, not a screenshot or a re-share. ### When is the person genuinely not findable? Some people simply are not on the public web in a way any face search can reach. A private social account, a closed group, a platform that blocks crawlers, or a person who has never posted a photo of themselves publicly all produce a legitimately empty result. Younger and more privacy-conscious users skew this way, and so do people in jurisdictions where platforms default to private. Absence of evidence here is genuinely weak evidence of absence. ### Should you search again with the same photo? Usually not. The index does change as pages are added, but re-running an identical image shortly after an empty result will almost always produce the same empty result, and it will spend credits doing it. A meaningfully different photo of the same person is worth far more than a repeat: a different angle, a different year, or a different lighting situation gives the model a different embedding to work with. If you only have one usable photo and it came back empty, the next useful step is a reverse image search instead, which asks a different question about the same file. Source: Face Search Returned No Results: What It Means and What to Try ## Face Search Glossary > Reverse face search borrows vocabulary from machine learning and from privacy law, and the two use some of the same words differently. These are the terms that come up when reading results or reading the rules, defined plainly, with a note on why each one matters in practice. Most misunderstood: - An embedding is a description of a face, not a picture of one, and cannot be turned back into a photo. - A confidence score measures similarity between two faces, not the probability that you are right. - A false positive is two different people scoring as a match, and it is the error that causes harm. - A biometric identifier is a legal category, and it is what most face-privacy law regulates. - Reverse image search and reverse face search are different operations on the same file. ### What is a face embedding? An embedding is the numeric description a model produces from a face: a fixed-length list of numbers capturing the geometric relationships between features. It is not a compressed photograph and cannot be reversed back into one. Two photos of the same person produce embeddings that sit close together in that numeric space, and two photos of different people produce embeddings that sit further apart. Everything a face search does is measuring that distance. ### What is a confidence score? A confidence score expresses how similar two faces are, on a scale the system defines. It is easy to misread as a probability that the match is correct, and it is not that. A score of 90 does not mean a 90 percent chance this is the same person. It means the two embeddings are very close by the model's measure, which is strong evidence in a good photo and much weaker evidence in a poor one. The photo quality is baked into how much a score is worth. ### What is a false positive? A false positive is when two different people are scored as a match. This is the error that does real damage, because it attaches a stranger's identity to someone. Lookalikes, siblings, and low-quality images all raise the rate. It is the reason a match should be treated as a lead to verify against other evidence rather than as a conclusion, and the reason no reverse face search engine can honestly claim to confirm identity. ### What is a false negative? A false negative is when the same person is present but not returned as a match. It is the quieter failure and the more common one in everyday use, caused by a bad angle, a filter, an obstruction, or simply a face that is not on any indexed public page. Because false negatives are common, an empty result is weak evidence that someone has no online presence. ### What is a biometric identifier? A biometric identifier is a legal category, not a technical one: data derived from a physical characteristic that can single out one individual. A face embedding usually falls inside it. This matters because most face-privacy law, including GDPR Article 9 and Illinois BIPA, regulates the processing of biometric identifiers rather than the act of looking at a photograph, so the rules can be triggered by computing the embedding at all. ### What is the difference between reverse image search and reverse face search? Reverse image search and reverse face search are two different operations on the same file. Reverse image search fingerprints the picture and finds copies of it: the same file reposted, cropped, or resized. Reverse face search discards the picture and keeps only a description of the face, then finds other photographs of that person, which may look nothing alike. Asking the wrong one is the usual reason a search seems to find nothing. ### What is liveness detection? Liveness detection decides whether a camera is seeing a real person present in front of it or a photograph, screen, or mask held up to it. It belongs to identity verification and unlocking, not to reverse face search. FaceSearch does not perform liveness detection, because it works from an image you supply rather than from a live camera. The distinction matters when reading privacy policies, since the two are regulated and discussed together but do very different things. ### What is an index in this context? The index is the searchable collection of face embeddings built from publicly reachable pages. A search compares your embedding against the index, so the index defines the boundary of what can possibly be found. Pages behind a login, blocked to crawlers, or never crawled are not in it, which is why private accounts are invisible regardless of how good your photo is. Source: Face Search Glossary: The Terms Explained Plainly ## How Does Facial Recognition Work? > Facial recognition converts a face into a fixed list of numbers called an embedding, then measures the distance between one embedding and another. Faces that produce nearby numbers are reported as the same person. Nothing in the process stores or compares the picture itself. In short: - A face becomes a vector of numbers; comparison is arithmetic on those numbers, not on pixels. - A match is a distance below a threshold, and the threshold is a choice, not a fact. - Lowering the threshold finds more true matches and more wrong ones, always both. - Searching a database of millions is a different task from checking two photos, and is measured separately. - The technology reads geometry. It does not know a name, and it cannot confirm one. ### How does facial recognition work, step by step? Facial recognition runs the same four steps regardless of which system performs it. The software differs enormously — a phone unlocking, a border gate, a face matching search on the open web — but the pipeline does not. Only the third step involves anything learned from data, where a trained algorithm turns the aligned face into numbers; the rest is preparation and arithmetic. - Detection: the system finds the region of the image that contains a face, and rejects images where it finds none. - Alignment: the face is rotated and scaled to a standard position so that two photos taken at different angles become comparable. - Embedding: a trained model converts the aligned face into a fixed list of numbers, typically between 128 and 512 of them. - Comparison: the distance between two embeddings is measured, and a distance below a chosen threshold is reported as a match. ### What is a face embedding, and what does it contain? An embedding is a list of numbers describing the geometry the model learned to treat as identifying — the relative proportions and structure of a face, not its pixels. It is not a compressed photograph, and a face cannot be read back out of it in any useful form. It is also not human-interpretable: no single number in the list corresponds to a nose or an eye. What makes it useful is only that two photos of the same person tend to produce embeddings that sit close together, and two different people tend not to. ### What decides whether two faces are called a match? A threshold does, and it is set by whoever operates the system rather than discovered in the data. Every threshold trades the two kinds of error against each other, and there is no setting that removes both. - A stricter threshold reports fewer matches and misses more real ones. - A looser threshold catches more real matches and returns more strangers. - The right setting depends on the cost of each mistake, which differs between unlocking a phone and searching the open web. ### Why is searching millions of faces harder than comparing two? They are different tasks and are measured separately. Comparing two photos is verification, written 1:1. Searching a large collection for a face is identification, written 1:N, and NIST evaluates the two in separate tracks for that reason. Every additional face in the collection is another opportunity for a stranger to fall inside the threshold, so an error rate that is negligible on one comparison stops being negligible when the comparison is repeated millions of times. ### What does facial recognition need from a photo? The model was trained on faces it could align, so the recoverable signal depends almost entirely on the input. Quality problems are not a small penalty; they usually decide the outcome. - The face should be large enough in frame that its structure survives compression. - It should face roughly toward the camera, because extreme angles hide the geometry the model relies on. - Lighting should be even, since a hard shadow across one side changes what the model measures. - Anything covering the eyes and the bridge of the nose removes the most informative region. ### How is facial recognition search different from facial recognition surveillance? They share the same mathematics and almost nothing else. A search compares one photo you supply against pages that are already public, and returns links to those pages. Surveillance systems match faces against an enrolled watchlist in a live camera feed, on a continuous basis, usually without the subject supplying anything. FaceSearch does the first and not the second: it takes a single uploaded image, searches public sources, and returns the pages where a similar face appears. It is not connected to any camera, and it holds no watchlist. ### What can facial recognition not tell you? It cannot tell you who someone is. It reports that two images are geometrically similar, and everything past that — the name, the account, the claim that both pictures show the same living person — comes from the page the match was found on, not from the model. A confident match on a page with the wrong caption is still the wrong answer, which is why a result is a starting point for checking rather than a conclusion. Source: How Facial Recognition Works: Embeddings, Thresholds, Limits ## How to Remove Your Face From Face Search Results > FaceSearch keeps no face index to remove you from: your upload is deleted within 24 hours and no face template is ever stored. Removal means acting where the photos actually live — the sites hosting them, the search engines listing them, and the face-search indexes that crawled them. The short version: - FaceSearch stores no biometric template and deletes uploads within 24 hours, so there is nothing held here to opt out of. - Photos are findable because they sit on publicly reachable pages. Removing the source is the only permanent fix. - FaceCheck.id, the index FaceSearch queries, runs its own removal process with identity verification. - Delisting from a search engine hides the result without deleting the page behind it. - In the EU, the UK and California, erasure is a right you can exercise against the site holding the photo. ### Does FaceSearch keep my face on file? No. The uploaded image is deleted within 24 hours of the search completing, and no biometric template, face print, or embedding is kept afterwards. The matching is transient: it happens during the search and nothing about your face persists once results are returned. There is no profile here to delete and no index entry to remove — which is also why FaceSearch cannot take a photo down on your behalf. ### Where do the photos actually come from? A face search does not hold pictures of you; it finds pages that already do. Search processing runs through the FaceCheck.id API, which matches against faces crawled from publicly reachable pages. That distinction decides your whole strategy: a result is a pointer, and deleting the pointer leaves the photo exactly where it was. - The hosting site is the source. While the photo is up, any index can find it again. - A face-search index is a copy of what it crawled. Removing yourself there stops that engine, not the others. - A general search engine is a third layer, with its own separate removal process. - Private and login-walled accounts are not crawlable, so they are not the ones surfacing. ### How do I remove my photos from FaceCheck.id? FaceCheck.id, the index behind FaceSearch results, publishes a removal page. You confirm the face is yours with a selfie or an anonymised identity document, and matching photos are hidden from new searches immediately while the check is processed. This is the fastest single action available, because it addresses the index FaceSearch actually queries. - Open the Remove My Photos page and search with a clear photo of yourself. - Select the images that are genuinely you. - Verify by selfie, which is usually automatic, or by identity document with the name and number obscured. - Approved photos are deleted, along with the document you uploaded. ### How do I get the result out of Google? Google runs a separate removal path for personal information, and it is worth being precise about what it does. Delisting removes the result from Search; the page it pointed to stays online and stays reachable by anyone with the address. It is a real reduction in exposure rather than a deletion, and it does not affect what a face-search index has already crawled. ### Can I make the website itself take the photo down? Often yes, and this is the only step that removes the photo rather than a reference to it. In the EU and the UK, the right to erasure lets you require a site to delete personal data it holds about you, and the operator has one calendar month to respond. California grants a comparable deletion right. The right is not absolute — it is weighed against public interest — but for an ordinary photo on an ordinary site it is usually the strongest lever you have. - Write to the site operator, name the photo and its URL, and cite your erasure right. - Keep a copy of the request and its date; the response clock starts there. - If the operator refuses or ignores you, your national data protection authority takes complaints. - Once the source is gone, ask the indexes to re-crawl or drop the stale copy. ### What if the photo is not mine to take down? Some images sit outside this process. A photo published by a news organisation, or one where someone else holds a genuine interest in it staying up, can survive an erasure request — the balance against public interest is real and it does not always fall your way. Intimate images taken or shared without consent are a different matter entirely, handled through dedicated reporting channels rather than a routine removal form, because most jurisdictions treat them as an offence rather than a privacy dispute. Source: How to Remove Your Face From Face Search Results ## How Accurate Is Reverse Face Search? > Reverse face search matches at approximately 95% accuracy under optimal conditions, meaning a sharp, front-facing, well-lit photo of one unobscured face. Accuracy is not a single number that holds everywhere. It falls with angle, blur, filters, and occlusion, and a confidence score measures similarity between two faces rather than the probability that a match is the right person. In short: - Approximately 95% is the figure under optimal conditions. It is not a guarantee that holds for every photo. - A confidence score measures how similar two faces are. It is not the probability that the match is correct. - False positives matter more than misses. Two different people scoring as a match is the error that causes harm. - Even an accurate matcher returns wrong people at web scale, because any index holds far more strangers than it holds the person being searched for. - Every match is a lead to verify against something else, never proof of identity. ### How accurate is reverse face search? Reverse face search reaches approximately 95% match accuracy under optimal conditions. Optimal conditions is a real constraint rather than a disclaimer: it describes one unobscured face, facing the camera, lit evenly, at enough resolution that the features are distinct. Move away from any of those and the accuracy of that individual search drops, even though the underlying model has not changed. The figure describes the system working on a good photo, not the average outcome of every photo a person happens to have. - The number describes the matcher, not your search. Your photo decides which end of the range you land on. - Accuracy and usefulness are different. A correct match on a page nobody can reach still tells you nothing. - There is no accuracy figure for a face that is not indexed anywhere. That is a coverage question, not an accuracy one. ### What does a confidence score actually measure? A confidence score measures how close the uploaded face and the matched face sit in the model's numeric space. It is a similarity measurement on the engine's own scale, not a probability that the two photos show the same person. A score of 90 does not mean there is a 90% chance the match is correct. It means the two faces are highly similar by the model's measure, which is strong evidence when the photo is good and much weaker when the photo is poor. ### Why does an accurate system still return the wrong person? Because any searchable index contains vastly more strangers than it contains the specific person being looked for. A matcher that wrongly pairs one face in ten thousand sounds precise, but a search comparing against millions of indexed faces will still surface wrong people alongside any correct one. Per-comparison accuracy does not survive being multiplied by the number of comparisons a web-scale search performs. This is the reason a confidence threshold exists, and the reason the top result is not automatically the right person. ### Which photo qualities change the result? Six properties of the uploaded image account for most of the variation between a clean match and an unusable result. They compound: a small, side-on, filtered face is far worse than any one of those problems alone. - Property Face angle. Helps accuracy: Front-facing, both eyes visible; Hurts accuracy: Steep profile or head turned away. - Property Resolution. Helps accuracy: The face fills much of the frame; Hurts accuracy: The face is a small crop from a wide shot. - Property Lighting. Helps accuracy: Even light across the whole face; Hurts accuracy: Hard shadow, backlight, strong colour cast. - Property Occlusion. Helps accuracy: Nothing covering the face; Hurts accuracy: Sunglasses, masks, hands, heavy hair. - Property Editing. Helps accuracy: The unedited original file; Hurts accuracy: Beauty filters, heavy retouching, AR lenses. - Property Subjects. Helps accuracy: One face in the photo; Hurts accuracy: Group shots, where the wrong face may be read. ### How is face recognition accuracy measured independently? The National Institute of Standards and Technology runs the Face Recognition Vendor Test, which evaluates submitted algorithms against controlled datasets and publishes the error rates. FRVT is the reference point for how good face recognition is in general, and it separates one-to-one verification, which confirms two photos show the same person, from one-to-many identification, which is what a face search performs. FaceSearch has not published its own benchmark against those datasets, and published FRVT rates come from curated images, so they describe an upper bound rather than what any web-scale search achieves in practice. ### When should a match not be relied on? A face match is evidence to check, and it is not sufficient on its own for any decision that affects a person. The error that causes harm is the false positive: a stranger who resembles the person being searched for, surfaced with a high score and acted on as though the identification were settled. - Do not act on a single match. Confirm it against something independent, such as a username, a mutual connection, or a second photo. - Do not use a match to decide someone's employment, housing, or credit. Those uses are regulated in several jurisdictions. - Do not read an empty result as proof that a person has no online presence. - Do not assume a low score means a different person. A low score often just means a poor photo. Source: How Accurate Is Reverse Face Search? ## How to Find Someone Using a Photo > To find someone from a photo, upload one clear image of their face to a reverse face search engine. The engine turns the face into a numeric descriptor, compares it against faces on publicly reachable pages, and returns the pages where a similar face appears. The search runs on the face itself, so no name, username, or caption is needed. In short: - One clear, front-facing photo of a single face is the whole input. There is no name field. - Crop to the face before searching if the photo is a group shot or a wide scene. - Results are pages where a similar face appears, ranked by similarity, each with a source link to open. - Only publicly reachable pages are searchable. Private accounts and closed groups are not. - A result is a lead. Confirm it against something independent before treating it as an identification. ### How do you find someone using a photo? Finding someone from a photo takes four steps, and the first one decides most of the outcome. The search compares facial geometry, so the effort belongs in choosing and preparing the image rather than in refining a query afterwards. - Choose a photo where the person faces the camera, the face is unobscured, and nothing is covering the eyes. - Crop to the head and shoulders if the photo contains several people, so the intended face is the one read. - Upload the image and run the search. FaceSearch accepts JPEG, PNG, and WebP files. - Open the ranked results and check each source page yourself, starting with the highest scores. ### Which photo gives the best chance of a match? The best photo for a face search is a sharp, evenly lit, front-facing shot of one person, saved from the original rather than re-shared. Screenshots of screenshots, heavy filters, and small crops from wide scenes all remove the fine detail the matcher depends on. If several photos of the person are available, running the clearest one first is more useful than running all of them, because each search consumes credits and a poor photo tends to produce poor results regardless of how many times it is submitted. ### What can a photo search find, and what can it not? A reverse face search returns pages on the public web where a matching face appears, which in practice means social profiles, dating profiles, public directories, news pages, and forum posts. It does not reach anything behind a login, and it does not read records of any kind. - It can find public social and dating profiles that use a different name from the one you know. - It can find the same person in photos taken years apart, since facial geometry is comparatively stable. - It cannot search private accounts, closed groups, direct messages, or anything requiring a login. - It cannot search by name, email address, or phone number, and it returns no government, medical, financial, or criminal records. - It does not notify the person searched for, and it does not build a profile of them. ### How is this different from a Google image search? A Google image search looks for the file. A reverse face search looks for the person. Uploading a photo to a general image search finds copies of that exact image and pages containing visually similar scenes, which works well for tracking where a picture has been republished. It will not find the same person in a different photo, because nothing in that pipeline models a face. The two answer different questions and are worth running in sequence rather than choosing between. ### What should you do with a match? Treat a match as a starting point and verify it before acting. Open the source page rather than judging from the thumbnail, then look for something that corroborates the identification independently of the face: a consistent username across sites, a mutual connection, a location or timeline that fits, or a second photo of the same person on the same profile. A high score with no corroboration is still a resemblance, and resemblance is exactly what a false positive looks like. ### Is it legal to search for someone by photo? Running a face search on a photo is lawful in most places, and what is done with the result is what the law tends to regulate. Several jurisdictions treat facial geometry as biometric data with its own consent rules, and using a match to decide employment, housing, or credit is restricted in more places than searching itself is. The rules differ enough by jurisdiction that the specifics matter. Source: How to Find Someone Using a Photo ## Reverse Video Search: Finding Someone From a Video > No face search engine searches a video file directly, and FaceSearch accepts JPEG, PNG, and WebP images only. To find a person who appears in a video, pause on a frame where the face is clear and forward-facing, save that frame as an image, and run a reverse face search on it. In short: - A video is not a searchable input. One frame from it is. - Pick a frame where the person faces the camera and is not moving, then crop to the head and shoulders. - Paused frames are usually softer than photos, so choose a still moment over a dramatic one. - Finding the person in a video and finding the source of the video are different tasks with different tools. - Frames from compressed or low-light footage often fail. Trying a second frame is cheaper than re-running the same one. ### Can you reverse search a video? Not as a video. There is no engine that accepts a video file and returns the people in it, and FaceSearch is no exception: its upload accepts JPEG, PNG, and WebP images only. What is described as reverse video search is in practice a two-part operation. A single frame is extracted, and that frame is then searched as an ordinary image. Everything that determines whether the search works is decided in choosing the frame. ### How do you get a searchable frame out of a video? Every platform gives some way to capture a still, and no special software is needed. The method matters less than the moment chosen. - Play the video and pause on a moment where the face is still, lit, and facing roughly toward the camera. - Step frame by frame if the player allows it. On many web players the comma and full stop keys move one frame at a time while paused. - Take a screenshot, then crop it down to the head and shoulders so the face is the largest thing in the image. - Save the crop as PNG rather than a re-compressed JPEG, which avoids adding a second round of compression artefacts. - Check the saved image at full size before uploading. If the eyes and mouth edges look mushy, choose a different frame. ### Which frame should you pick? The best frame is a still moment, not an interesting one. Motion is the enemy of a face match, because a moving face is blurred across the exposure and the fine geometry the matcher depends on is smeared away. A person listening is a better frame than the same person mid-sentence. - Prefer A pause or a held expression. Avoid: Mid-gesture or mid-word; Why: Motion blur removes the feature detail entirely. - Prefer Face turned toward the camera. Avoid: Profile or a glance away; Why: A profile hides half the geometry being compared. - Prefer Steady, even light. Avoid: Backlight or a strobing scene; Why: Silhouettes and colour casts flatten the features. - Prefer A close or zoomed shot. Avoid: A wide or crowd shot; Why: A small face survives compression badly. - Prefer Original upload quality. Avoid: A screen recording of a playing video; Why: Each re-encode compounds the artefacts. ### Why are video frames harder to match than photos? A video frame is a worse photograph than a photograph. Video compression allocates detail to what changes between frames rather than to the sharpness of any single one, so a paused frame carries visible blocking and smearing that a still image of the same scene would not have. Rolling shutter can skew a moving face, low-light footage is noisier than a low-light photo because the exposure per frame is short, and any re-upload compresses the whole thing again. This is why the same person can match easily from a profile picture and fail from a frame of the video that profile picture was taken from. ### How do you find the source of a video rather than a person in it? Finding where a video came from is a different task and needs a different tool. A face search answers who is in the frame, so it does not help with an unattributed clip, a re-uploaded stream, or a suspected deepfake's origin. For provenance, take a keyframe or the thumbnail and run it through a general reverse image search, which looks for copies of the file and visually similar scenes across the web. Run the two in sequence when both questions matter: the image search for where the footage has been published, the face search for who appears in it. ### What about live streams and screen recordings? The same rule applies, with one extra loss of quality to plan around. A screen recording re-encodes footage that was already compressed for delivery, so a frame taken from a recording of a stream is two generations away from the original and usually the weakest input available. Where the platform offers it, capture a still directly rather than recording the screen and extracting a frame afterwards, and prefer a moment when the stream was not dropping quality to keep up. Source: Reverse Video Search: Find Someone From a Video ## How to Verify Someone You Met Online > Verifying someone you met online means establishing that the person behind the profile is the person they claim to be, whether you met on a dating app, a marketplace, or a group chat. No single check does that. A reverse face search is the fastest first step, because it shows where else that face has been published and under what name, and the rest is cross-checking what they told you against sources they did not choose. In short: - Verification is a chain of small checks. No single result, a face match included, proves that someone is who they say they are. - A reverse face search tells you where a photo was published before, and under which name. - An empty result is weak evidence. Private accounts are invisible to every face search. - A live video call is still the strongest check available to a private individual. - Checking someone you deal with personally is not the same as screening them, and screening is not a permitted use. ### How do you verify someone you met online? Work outward from the photo they chose towards sources they did not. The photo is the only thing you can check without asking them for anything, which is why it comes first; everything after it is about finding one fact they did not supply themselves. - Run a reverse face search on their clearest profile picture, and read the source pages rather than the scores. - Check whether the name on those pages is the name they gave you. - Check how far back the matching pages go. A face that first appears online this year is a different signal from one with a decade of history. - Ask for a live video call, and ask them to do something specific during it. - Verify one concrete detail — an employer, a city, a university — against a source they did not send you. - Treat money, secrecy or urgency as a reason to stop, not a detail to work around. ### What does a face match actually prove? It proves that a visually similar face has been published on a page the index could reach. It does not prove identity and it cannot: every name attached to a matching page was written by whoever published that page, and verified by nobody. A high confidence score means two images are geometrically close, not that the caption above them is true. Read the source page before you act on the match. ### What should you check when a face search finds nothing? Change what you are checking rather than repeating the same search. Most people's photos sit behind private accounts that no face search reaches, so an empty result says little about the person and nothing about their honesty. - Ask for a live video call rather than more photographs. - Search the exact sentences from their profile. Scam scripts are reused word for word across accounts. - Check whether their posting history matches the age and life stage they describe. - Ask something only the real person would be able to answer without looking it up. ### How do you verify someone before meeting them in person? Do the identity checks and the safety arrangements separately, because passing one does not remove the need for the other. The Federal Trade Commission's guidance for people meeting someone from online is deliberately practical, and its strongest single rule is about money rather than identity: never send money or gifts to someone you have not met. - Video call before you agree to meet, not after. - Meet in a public place, in daylight, the first time. - Arrange your own transport there and back. - Tell one person where you are going and when you expect to be back. ### Which questions can a face search not answer? It cannot tell you whether someone is dangerous, solvent, employed, or truthful about anything other than the photograph itself. It is also not a screening tool: decisions about employment, housing, credit, insurance or education are outside what FaceSearch permits, whoever the subject is. Verifying a person you are personally dealing with and screening a person on someone else's behalf are different acts, and only the first is in scope. ### When have you verified enough? When the account holds up against at least one source the other person did not choose, and nothing you found contradicts what they told you. That is a lower bar than proof and a higher one than a good feeling. If you cannot reach it — no matching pages, no video call, no independently checkable detail — the honest conclusion is that you have not verified them, and that is itself a result worth acting on. Source: How to Verify Someone You Met Online: A Step-by-Step Check ## How to Run a Reverse Face Search on a Phone > A phone can run a reverse face search in full: open FaceSearch in the mobile browser or the app, choose a photo from the camera roll or a screenshot, crop to the face, and search. The steps are the same on iPhone and Android; the differences are in how each handles photo formats and cropping. In short: - No desktop is needed. The whole procedure works from the camera roll. - Crop to the head and shoulders before searching, especially on a screenshot. - Screenshots of a playing video work, but pause first so the frame is not blurred. - iPhone photos may be HEIC; convert or screenshot them, since the search accepts JPEG, PNG and WebP. - Portrait mode blurs the background, not the face, so it is safe to search. ### Can you run a reverse face search from a phone? Yes, and nothing about the search is reduced on mobile. The same sources are searched and the same results are returned as on a desktop. FaceSearch runs in any mobile browser and also ships native apps for iPhone and Android, so a photo can go from the camera roll to a result without leaving the phone. ### How do you search a photo that is already on your phone? The camera roll is the usual starting point, and the only preparation that matters is making sure the intended face is the one being read. - Open FaceSearch in the browser or the app and choose the upload control. - Pick the photo from the camera roll, or take one now. - If more than one person is in the shot, crop to the head and shoulders of the person you mean before uploading. - Run the search and open each result at its source page rather than judging from the thumbnail. ### How do you search a face from a screenshot? A screenshot is often the only way to get an image out of an app that does not allow saving, and it works well as long as the capture is deliberate. Take the screenshot, then crop away the interface — the status bar, buttons and captions are not part of the face and only shrink the region the model gets. If the screenshot is of a video, pause it first: a frame captured mid-motion carries blur that the search cannot recover from. ### Does the iPhone photo format matter? It can. iPhones save photos as HEIC by default, and FaceSearch accepts JPEG, PNG and WebP. Most sharing paths convert automatically, so this rarely surfaces, but if an upload is rejected there are two reliable fixes: take a screenshot of the photo, which produces a PNG, or switch the camera to Most Compatible in Settings so new photos are saved as JPEG. Portrait mode needs no special handling — it blurs the background and leaves the face sharp. ### Should you use the app or the mobile site? Either. The apps and the site search the same index and return the same results, and an account carries its credits and history across all three. The apps are more convenient for repeated searches because the camera roll and the camera are one tap away; the mobile site is the faster choice for a photo someone has just sent you in a browser. ### Why does a phone search sometimes return fewer matches? Almost always because of what happened to the image before it was uploaded, not because of the phone. Every app that forwards a photo re-compresses it, and each pass removes detail the model needs. - A photo saved from a chat app has usually been compressed at least twice. - A screenshot of a photo inside another app is a copy of a copy. - Zooming in before screenshotting enlarges pixels without adding detail. - Where possible, get the original file rather than a forwarded version of it. Source: Reverse Face Search on a Phone: iPhone and Android ## How Much Does a Reverse Face Search Cost? > One reverse face search on FaceSearch costs 3 credits. A new account starts with 3 free credits, which is one complete search, and no credit card is required to use them. Further credits come either from a monthly plan or from a one-time pack, and the price per credit falls as the pack size rises. In short: - A search costs 3 credits, and a new account gets 3 free credits, which is exactly one search. - No credit card is needed to run that first search. - The monthly plan is cheaper per search than any pack below the largest ones, and it renews. - One-time packs never expire into a subscription. They suit occasional use. - Credits are consumed by the search, not by the result, so a search that finds nothing still costs credits. ### How much does a reverse face search cost? A single reverse face search on FaceSearch costs 3 credits, and every account begins with 3 free credits, so the first search costs nothing and requires no payment details. After that, credits are bought either through the monthly plan, which grants a fresh allowance every month, or through a one-time pack, which is a single purchase with no renewal. The table above converts both into a cost per search, which is the only figure worth comparing across the two. ### Is there a genuinely free reverse face search? Free reverse face search generally means a free first search rather than an unlimited free service, and FaceSearch works the same way: 3 credits on signup, one complete search, no card. The reason the category has no permanently free tier is that each search is a comparison against a large index of faces, which costs compute every time it runs, unlike a keyword search that can be served from a cache. A service advertising unlimited free face search is usually funding it another way, and the uploaded photo is the most likely source of that funding, so it is worth reading what a free provider says it retains before uploading a face to it. ### What is a credit, and why does one search cost several? A credit is the unit FaceSearch bills in, and one search consumes 3 of them. Pricing in credits rather than in searches is what keeps a plan and a pack directly comparable, since both are denominated in the same unit. Credits do not expire while an account is open, and credits bought in a pack sit alongside any granted by a plan rather than replacing them. ### Should you choose the monthly plan or a credit pack? Choose the monthly plan for recurring work and a pack for a single question. The plan renews its allowance each month and costs less per search than the smaller packs, which suits anyone verifying profiles regularly. A pack is a one-time purchase that does not renew and does not need cancelling, which suits someone with one person to check. The largest packs reach a lower per-credit price than the plan, so high-volume one-time use is better served by a large pack than by a subscription. ### Do credits get used up when a search finds nothing? Yes. Credits pay for the search being run, not for the result it returns, so a search that finds no matches consumes the same 3 credits as one that finds many. This is worth knowing before spending credits on a poor photo: the comparison against the index happens either way, and an unclear or heavily filtered image is the most common reason a search completes with nothing to show. Improving the photo before searching is cheaper than searching twice. ### How do you pay, and can you get a refund? Payment goes through Stripe for card transactions, with cryptocurrency accepted separately, and FaceSearch does not store card details on its own servers. Refunds are governed by the published refund policy rather than decided case by case, and because credits are consumed at the moment a search runs, the policy is the thing to read before buying rather than after. Source: How Much Does a Reverse Face Search Cost?