Username Checker & Verified Profile Finder

Search a username across hundreds of public sites and return confidence-filtered profile URLs. Built as a username checker, OSINT username search, social profile finder, and account-discovery API for AI agents.

Built for AI agents, OSINT workflows, trust & safety tooling, account discovery, brand monitoring, and username research.

What this Actor does

Give the Actor one or more usernames. It checks public websites using Maigret and returns structured JSON with confidence levels.

Unlike basic username-search wrappers, this Actor does not treat every HTTP hit as equally trustworthy. Search pages, generic pages, HTTP errors, and weaker detection methods are downgraded instead of being presented as verified profiles.

Key features

• Agent-friendly structured JSON
• Fast mode: checks the top 200 sites
• Deep mode: checks the top 500 sites
• High / medium / weak confidence classification
• Strict mode for high-confidence results only
• No paid upstream API dependency
• Pay-per-event pricing
• Works well from API, automations, and AI agent workflows

Input

{
  "usernames": ["openai"],
  "mode": "fast",
  "strict": true,
  "timeoutSeconds": 5
}

Input fields

Field	Type	Description
usernames	array	1-20 usernames to search
mode	string	fast checks top 200 sites, deep checks top 500
strict	boolean	When true, profiles contains only high-confidence matches
timeoutSeconds	integer	Per-site network timeout, 3-15 seconds

Output

Each username produces one dataset item.

{
  "username": "openai",
  "ok": true,
  "mode": "fast",
  "sitesChecked": 200,
  "elapsedSeconds": 12.1,
  "verifiedProfileCount": 17,
  "highConfidenceCount": 17,
  "mediumCandidateCount": 56,
  "weakSignalCount": 29,
  "profiles": [
    {
      "site": "YouTube",
      "url": "https://www.youtube.com/@openai/about",
      "confidence": "high",
      "checkType": "message",
      "httpStatus": 200,
      "tags": ["social", "streaming", "video"]
    }
  ]
}

Confidence model

high means the site returned a direct-profile-style URL and Maigret detected explicit presence/absence markers with no known protection signal.

medium means the profile is a plausible candidate but the evidence is weaker. Medium candidates are not billed as verified profiles.

weak includes HTTP-only signals, search/filter pages, generic roots, HTTP errors, and other low-confidence detections.

Confidence refers to technical evidence that a public username profile exists. It does not prove that accounts using the same username belong to the same real-world person.

Pricing

This Actor uses Pay Per Event pricing.

• Username Checked: $0.01 per successfully completed username search
• Verified Profile Found: $0.012 per high-confidence profile returned
• Actor Start: $0.00005 per run
• Platform usage is included for the customer

Failed username searches are not charged by the custom username/profile events.

Good use cases

• Username search API
• Public social profile finder
• OSINT enrichment
• Account discovery for AI agents
• Brand or handle monitoring
• Trust & safety research
• Investigating public username reuse
• Building MCP / agent workflows that need structured profile evidence

Important limitations

This Actor searches public information only. A matching username across multiple websites does not establish common ownership or identity. Results may change as websites change their HTML, anti-bot systems, APIs, or profile rules.

Do not use the output as proof that multiple profiles belong to the same person. Treat results as public username-presence evidence and candidate profiles that may require independent verification.

Responsible use

Use this Actor only for lawful purposes involving public information. Do not use it to harass, stalk, doxx, impersonate, or target individuals.

Technology and license

The search engine is based on Maigret 0.6.6, which is MIT licensed. This Actor adds filtering, confidence classification, billing logic, Apify packaging, and agent-oriented structured output.