Facial Search API

Facial Search API for Identity Verification

Submit a face image and receive matched online profiles, linked names, web presence, and identity consistency signals — all in a single structured API call. Built for KYC, fraud prevention, OSINT, and risk enrichment workflows that start with a photo.

Match a face image to online profiles, social accounts, and web presence
Impersonation and identity consistency checks with confidence scoring
Returns linked names, URLs, and matched profile sources per result
Accepts image URL or base64 — JPEG or PNG
FR
Facial Search
ClearCheck Data Engine
Profiles Found
Image Input
Face detected & processing
8 profiles matched
LinkedIn
linkedin.com/in/…
97%
Twitter / X
twitter.com/…
91%
Instagram
instagram.com/…
84%
Impersonation Risk
No flags detected
Clear
Best match confidence97/100

From a Face Image to a Full Identity Signal Set

Submit a face image URL or base64 payload and receive matched profiles across the web with confidence scores, linked names and URLs, impersonation risk flags, and an overall identity consistency score — in a single API response.

Image Matching

Search the web for publicly available images visually similar to the submitted face. Returns matched profiles ranked by confidence score — from highest probability matches down to lower-confidence associations.

Profile Discovery

For each matched face, returns the associated profile URLs, platform names, linked usernames, and public page information. Surfaces the breadth of a person's digital presence from a single image input.

Impersonation Check

Detects cases where the same face image appears under multiple conflicting identities — a strong indicator of account impersonation, identity fraud, or synthetic profile creation across platforms.

Web Presence

Returns all public web pages and domains where the matched face has appeared. Builds a complete map of the subject's online footprint — from professional profiles to social media to news appearances.

Identity Consistency

Scores how consistently the matched face appears under the same name and identity across sources. Low consistency scores indicate conflicting identities, possible impersonation, or use of a stolen photo.

Confidence Score

Each match is assigned a confidence score between 0 and 1 representing the visual similarity between the submitted image and the matched profile photo. Returns best_confidence for the top result and per-match scores for all results.

What You Send

Submit the face image as a URL or base64 payload. Only the image and API key are required — the lookup ID scopes optional result filtering.

Face Image

A direct URL to a publicly accessible JPEG or PNG image, or a base64-encoded image payload. The image should clearly show a face. Minimum recommended resolution is 128×128 pixels.

API Key

Your ClearCheck Data API key is passed in the request body. Always kept server-side and never exposed in frontend code or client-side JavaScript.

Lookup ID

The lookupId parameter scopes the search to a specific data source group. Refer to the API documentation for available lookup IDs and their coverage scope.

Responsible Use Notice: Facial search and image-based identity enrichment must only be used for lawful purposes — including fraud detection, KYC, identity verification, and legitimate investigative workflows. Use must comply with applicable biometric data privacy laws, GDPR, CCPA, and any applicable industry regulations. Customers are solely responsible for ensuring lawful and ethical use of this API.

What You Get Back

Every completed facial search returns a structured JSON record with all matched profiles, confidence scores, linked identities, and impersonation flags.

match_countTotal number of unique profile matches returned for the submitted face image across all searched sources.
best_confidenceThe highest confidence score among all returned matches. Range 0–1. Scores above 0.90 indicate high-confidence visual similarity.
matched_profiles[]Array of matched profiles, each with profile URL, platform, linked names, confidence score, and first/last seen dates.
linked_names[]All names found associated with the matched face image across returned profiles. Multiple conflicting names can indicate impersonation.
linked_urls[]All web URLs where the matched face has appeared, across social media, professional networks, news, and other public web sources.
impersonationBoolean flag indicating whether the same face was found under multiple conflicting identities — a common pattern in account fraud and identity theft.
web_presenceSummary of the subject's public web footprint — number of sources, earliest and most recent appearance, and primary platforms where the face was found.
consistency_scoreA 0–1 score reflecting how consistently the face appears under the same identity. Low scores flag possible impersonation or stolen photo use.
identity_scoreComposite identity signal score combining match confidence, consistency, and impersonation flags into a single actionable enrichment metric.

Submit, Receive, Poll, Parse

The facial search runs through an async enrichment pipeline. Submit the image, hold the job ID, and poll when ready — your backend keeps processing without waiting on the search to complete.

1

Submit Face Image

POST an image URL or base64 payload with your API key and lookup_type "face". The image should be a clear JPEG or PNG of a face.

2

Receive Job ID

The API responds immediately with a numeric job ID and status "progress". No waiting required — your workflow continues immediately.

3

Poll for Completion

Query the monitor endpoint using the job ID. When status changes to "completed", all matched profiles and confidence scores are ready.

4

Use the Results

Parse matched_profiles for profile URLs, check impersonation flag, read consistency_score for identity analysis, and feed linked_names to name lookup for further enrichment.

How to Make a Facial Search Call

Send one POST request with the face image URL and your API key. Get a job ID back immediately — no waiting. Poll once to pick up the full set of matched profiles when the search is complete.

Set lookup_type to face and pass the image either as an image_url pointing to a publicly accessible JPEG or PNG, or as a base64 payload in image_base64.

API key in request body — never exposed in frontend code
Accepts both image URL and base64 payload — JPEG or PNG
Non-blocking — submit and continue, collect results when ready
Per-match confidence scores plus impersonation and consistency signals

Docs are currently on our existing portal and will move to the ClearCheck Data portal soon.

POST /api/developer/combined_face JSON
// 1. Submit facial search
POST /api/developer/combined_face
Content-Type: application/json
{
  "key":          "YOUR_API_KEY",
  "lookup_type": "face",
  "image_url":   "https://example.com/photo.jpg",
  "lookupId":   93
}

// 2. Initial response
{
  "id":     7738204,
  "status": "progress"
}

// 3. Poll for result
GET /api/request-monitor/api-usage/7738204

// 4. Completed response
{
  "match_count":    8,
  "best_confidence": 0.97,
  "impersonation":  false,
  "identity_score": 0.91,
  "matched_profiles": [...]
}

Who Uses Facial Search and Why

Facial search is a high-value enrichment signal for identity verification, fraud prevention, and investigation workflows where a photo is the primary available identifier.

KYC & Onboarding

During document-based KYC, compare the applicant's submitted photo against their online presence. A face that doesn't match any public records under the claimed name is a strong identity risk signal.

Fraud Prevention

Detect stolen photo fraud — where bad actors use someone else's image as their profile photo. Impersonation detection flags faces found under multiple conflicting identities for immediate investigation.

OSINT & Investigations

Start an investigation with just a photo. Facial search maps the subject's online presence, surfaces linked usernames and profile URLs, and builds an identity picture for further enrichment with Name or Email Lookup.

E-commerce Verification

For high-value or high-risk transactions, a facial search of the submitted account photo adds an extra identity confidence layer — confirming the user is who they claim to be against their public online identity.

Risk Orchestration

Integrate facial search results into your broader risk scoring pipeline. Feed identity_score and impersonation flag into your decision engine to automatically route applicants based on biometric identity confidence.

Marketing & CRM

For legitimate outreach and influencer enrichment use cases, facial search surfaces linked social profiles from a known photo — helping marketing teams understand a contact's full social media presence from a single image.

Frequently Asked Questions

The API accepts JPEG and PNG images submitted either as a publicly accessible image URL (via the image_url field) or as a base64-encoded payload (via image_base64). The image should clearly show a face. While the API will attempt processing on smaller images, a minimum recommended resolution of 128×128 pixels ensures the best matching results. Multi-person images may return multiple match sets.
The Facial Search API searches across publicly available web sources — including social media platforms, professional networks, news sources, and other publicly indexed pages where profile photos and images are present. The specific coverage depends on the lookupId parameter used. Refer to the API documentation for the available lookup configurations and their respective data source coverage.
Confidence scores range from 0 to 1 and represent visual similarity between the submitted image and the matched profile photo. A score of 0.91 means there is a 91% visual similarity — a high-confidence match. Scores above 0.90 generally indicate the same person. Scores between 0.70 and 0.89 are moderate-confidence matches worth reviewing manually. Scores below 0.70 are low-confidence and may represent partial or incidental similarity.
Impersonation is detected when the same face is found appearing under multiple conflicting names or identities across the searched sources. For example, if a face appears as "Alex Johnson" on LinkedIn but as "Maria Torres" on a different platform, the impersonation flag is set to true. This pattern is common in synthetic identity fraud, account takeover, and social engineering attacks. The consistency_score reflects the degree of name and identity consistency across all matched profiles.
Facial search runs as an async operation. After submitting the image you receive a job ID immediately. Search processing typically completes within 30 to 120 seconds depending on image quality, the number of sources searched, and current system load. Poll the monitor endpoint with your job ID and proceed when status changes to "completed". You can implement exponential backoff on polling to avoid unnecessary requests.
Biometric data, including facial imagery, is subject to strict legal requirements in many jurisdictions — including GDPR in the EU, BIPA in Illinois, CCPA in California, and other applicable biometric privacy laws. Use of the Facial Search API must be limited to lawful purposes such as fraud detection, KYC, identity verification, and authorized investigative workflows. Customers are solely responsible for ensuring their use of this API complies with all applicable laws, their own privacy policies, and applicable industry regulations.