Social Profile Finder

Find public profiles for a username across more than 400 social networks, developer communities, forums, media sites, and other websites (404 sites in the pinned checker definition set). Use it for OSINT research, brand protection, lead enrichment, identity resolution, and due diligence.

Best first run: start with one username and three exact sites, such as GitHub, Reddit, and YouTube. Review the candidates and diagnostics before expanding to the full site list. Open Social Profile Finder.

Each username produces one structured dataset row, making the results easy to export to JSON, CSV, Excel, Make, Zapier, or your own API workflow.

Every run also writes RUN_SUMMARY with completed searches, failed searches, partial searches, and total accounts found. If every checker process fails, the run is marked FAILED after the diagnostic rows are saved; an empty result from a completed search is kept distinct from a checker failure.

Every dataset row now records search_mode and sites_requested, so a focused three-site result cannot be mistaken for a broad 400+ site search. Individual matches include evidence: "sherlock_explicit_claim" and verification_required: true; they are candidates to verify at the source, not identity guarantees.

Best for

• Qualifying a username across a focused set of networks before a broader search
• Enriching lead lists with public profile URLs and explicit checker diagnostics
• Investigating possible impersonation or brand presence without treating an empty result as proof of absence

What you get

• Profile URLs grouped by username and site
• Targeted searches on selected sites for faster results
• Broad discovery across 400+ supported sites when no filter is supplied
• Batch input for up to 20 usernames per run
• A completion flag and error message for every username, including partial failures
• No third-party API key required

Example input

Use the focused example below when you want a fast qualification check. Site names must match the supported names exactly; leave sites empty when you want broad discovery. The form does not preselect any sites, so a first run checks every supported site unless you add your own list. Site names the checker does not recognize are reported in error_message and in sites_invalid instead of being silently skipped.

For a fast, focused search:

{
  "usernames": ["openai"],
  "sites": ["GitHub", "Reddit", "YouTube"]
}

You can paste handles with or without a leading @; the Actor normalizes them before searching. API calls follow the same 20-username limit as the Store form.

To check every supported site, omit sites:

{
  "usernames": ["openai", "github"]
}

The Store form intentionally leaves sites empty so a new run is not silently limited to a repeated shortlist. Add exact Sherlock site names when you want a faster, focused qualification pass.

Pricing

This Actor uses pay-per-result pricing: each saved username search result is one paid result, including a completed search with no matching profiles. One result can contain multiple candidate profiles; you are not charged separately for every profile in that result. The current fee is about $4.99 per 1,000 results, plus $0.00005 per Actor start event (one event per GB of memory, minimum one). Apify platform usage is included. Check the Store Pricing tab for current rates. Set a maximum run charge to control spending, and start with a focused site list before expanding your search.

Ready-to-run presets

Broad discovery:

{
  "usernames": ["openai"],
  "sites": [],
  "maxConcurrency": 2
}

Focused qualification:

{
  "usernames": ["openai"],
  "sites": ["GitHub", "Reddit", "YouTube"],
  "maxConcurrency": 2
}

Batch brand check:

{
  "usernames": ["openai", "example-brand"],
  "sites": ["GitHub", "YouTube", "LinkedIn"],
  "maxConcurrency": 2
}

Start with focused mode when qualifying a lead, then run broad discovery only for usernames that need wider coverage.

Input

Field	Type	Required	Description
usernames	string[]	Yes	One to 20 usernames to search.
sites	string[]	No	Exact supported site names, case-sensitive. Empty means every supported site.
maxConcurrency	integer	No	Parallel username searches, from 1 to 4. Default: 2.

Selecting a short list of sites reduces runtime and platform compute cost. A broad search takes longer because it checks hundreds of independent websites.

Output

Results are saved to the run's default dataset. One input username produces one row:

{
  "username": "openai",
  "search_mode": "focused",
  "sites_requested": ["GitHub", "Reddit", "YouTube"],
  "result_interpretation": "candidate_profiles_verify_at_source",
  "total_found": 2,
  "search_completed": true,
  "partial_failure": false,
  "sites_invalid": [],
  "error_message": null,
  "accounts_found": [
    {
      "site": "GitHub",
      "url": "https://www.github.com/openai",
      "username": "openai",
      "status": "found",
      "evidence": "sherlock_explicit_claim",
      "verification_required": true
    },
    {
      "site": "Reddit",
      "url": "https://www.reddit.com/user/openai",
      "username": "openai",
      "status": "found",
      "evidence": "sherlock_explicit_claim",
      "verification_required": true
    }
  ]
}

If partial_failure is true, the row contains only Sherlock's explicit claimed profile URLs but at least one checker reported an error. Treat those URLs as leads, inspect error_message, and verify important matches at the source before using them. A search_completed: false row has no trustworthy profile claims and should be retried. The run summary exposes the count as searchesPartial.

Live validation

A focused build 0.1.37 canary on 2026-08-15 checked openai on GitHub, Reddit, and YouTube. It completed successfully, returned the three expected public profile URLs, recorded the three-site filter explicitly, marked every claim as requiring source verification, and used $0.00391 in platform resources.

You can retrieve the same data through the Apify API:

GET https://api.apify.com/v2/datasets/DATASET_ID/items?clean=true&format=json

Common workflows

• Enrich a lead list with likely social profiles
• Investigate a public username during OSINT research
• Detect possible brand or executive impersonation
• Map a creator's public presence across platforms
• Feed profile URLs into a downstream scraper or review queue

Continue the workflow

• LinkedIn Ad Library Scraper — add LinkedIn campaign and advertiser research to a qualified brand profile.
• Google Ads Transparency Scraper — compare a brand's public ad creatives after finding its official profiles.
• YouTube Ads Scraper — collect matching YouTube ad videos and landing pages for creative research.

Use the focused sites input first for a quick qualification pass, then expand to the full supported-site list only when you need broader discovery.

Accuracy and responsible use

The Actor checks publicly reachable profile URLs. Websites can change their responses, block automated requests, or return ambiguous status pages, so results may contain false positives or miss existing profiles. Treat matches as leads and verify important findings at the source.

Use the Actor only for lawful purposes and handle personal data in accordance with the rules that apply to your use case. The Actor does not access private accounts or bypass authentication.

Troubleshooting

• A site name is rejected: use the exact supported site name, including capitalization. Unsupported names are listed in error_message and in the row's sites_invalid field, and any other requested sites are still checked.
• A broad run is slow: select only the sites relevant to your workflow.
• A username returns no matches: verify the spelling and retry with a focused site list.
• search_completed is false: check error_message, then retry that username in a smaller run.
• partial_failure is true: use the explicit profile claims as leads, read error_message, and verify the URLs at their source.
• A checker or site-definition download failure is recorded as search_completed: false; do not treat that row as a confirmed no-match.

Billing and spending limits

Completed username searches, including completed searches with no matches, are paid results. Partial searches with candidate profiles are also paid results. Failed checker searches and partial searches without useful profiles are saved to the separate searchDiagnostics dataset without a result fee. The Actor stops starting searches when its result budget is exhausted. See RUN_SUMMARY for saved results, unbilled diagnostics and skipped usernames.

Recently fixed

• The Store form no longer preselects GitHub, Reddit, and YouTube, so a first run really covers every supported site instead of a silent three-site shortlist.
• Unsupported site names are now reported in error_message and sites_invalid instead of being dropped without a trace, and an all-invalid filter fails with the offending names rather than a generic retry message.