Username Search Tools Compared: Sherlock, Maigret, WhatsMyName and Platforms

Disclaimer: This article is for education and for lawful, authorised professional research. Use these methods only where you have a legitimate purpose and a lawful basis, and follow the laws and platform terms that apply to you, including data protection law such as the UK GDPR and EU GDPR. See our Terms of Service.

TL;DR

• Sherlock vs Maigret vs WhatsMyName is the usual comparison, but all three, plus Blackbird and Social Analyzer, answer one question: does an account with this username exist on site X? They differ in site coverage, how they decide "found", how they handle false positives, and how actively they are maintained.
• Maigret has the largest database and extracts linked identifiers for recursive searches. WhatsMyName has the most careful detection data. Sherlock is the simplest to run. Blackbird is a WhatsMyName-based checker. Social Analyzer's repository is quiet and its PyPI package has not changed since 2022.
• Every tool produces false positives and false negatives. Treat a "found" as a lead to open and verify, never as a finding.
• We checked every repository's last commit and release on 24 September 2026. Dates are in the comparison table.
• An aggregated platform such as UserSearch Username Intelligence, with OneScan and Cases, fits when you need several sources, source attribution and an audit trail in one place.

Five Tools, Five Different Answers for One Handle

Sherlock vs Maigret vs WhatsMyName comes down to trade-offs: Maigret has the largest database and recursive search, WhatsMyName has the most careful detection data, and Sherlock is the simplest to run. Run the same username through three open-source checkers and you will get three different lists. One reports 41 accounts, another 27, a third 63. Some sites appear in all three, some in one. A few of the hits are plainly wrong: a site that returns a normal page for any username, or a login wall that the tool reads as an account page.

None of that means the tools are bad. It means they make different trade-offs, and an analyst who does not know those trade-offs will either miss accounts or report ones that do not exist. This comparison is for analysts choosing a tool, or a combination, for username research in due diligence, fraud prevention, security research and journalism. We cover what each tool does, how to install and run it, how it behaves, and how actively it is maintained, with dates.

What Does a Username Checker Actually Do?

A username checker is a loop. For each site in its database, it builds the URL where an account page would live, such as https://example.com/user/{username}, sends a request, and reads the response to decide whether the account exists. The database entry tells it how to decide. The WhatsMyName project shows this clearly, because the data is all it publishes. Its JSON file is open on GitHub and powers the free whatsmyname.app web tool. Each site entry includes:

• uri_check: the URL template, with {account} where the username goes.
• e_code and e_string: the HTTP status code and a text string expected when the account exists.
• m_code and m_string: the code and string expected when it does not.
• known: usernames known to exist on the site, used to test that the entry still works.

The quality of a checker comes down to three things: how many sites are in the database, how precise each detection rule is, and how quickly broken rules are fixed when a site changes its pages. A rule that relies only on the status code will misfire on the many sites that return 200 OK for everything. A rule that checks for a specific string on the page and a different string on the "not found" page is far more reliable.

Why Does Tool Choice Change Your Findings?

Username reuse is one of the most productive links in open-source research, and recent public work shows both its power and its limits. Bellingcat's researchers made the same point in their April 2024 guide to common mistakes in open-source research: a shared name or a similar-looking photo is not enough on its own to link two things, and several public misidentifications happened because people treated it as enough. A username match is the same kind of evidence. Treat it as confirmed only after independent details line up.

In research like that, the account list is only the starting point, and missing one platform, or trusting one false hit, can change the result. That is why the choice of tool matters, and why many teams run more than one. If you do this work professionally, UserSearch 2.0 is built for exactly this: many sources, one account, results you can check.

Installing and Running Each Tool

All five need Python, apart from Social Analyzer's Node.js option. We recommend installing each in its own virtual environment or with pipx, because their dependencies conflict with each other more often than you would expect. We use example_handle as the username throughout.

Sherlock: The Simple Default

Sherlock is the best-known checker and still the quickest to get running.

pipx install sherlock-project
sherlock example_handle --print-found --csv --timeout 20

• pipx install sherlock-project installs the CLI in its own environment. The package needs Python 3.9 or later. Docker (docker run -it --rm sherlock/sherlock) also works. The project warns that some third-party distribution packages have been broken, so prefer pipx, pip or Docker.
• --print-found shows only hits, --csv writes a spreadsheet next to the default text file, and --timeout 20 cuts the per-request wait from its default of 60 seconds, which otherwise lets one slow site hold up the run.
• --site limits the run to named sites, and {?} in a username checks variants with _, - and ..

Sherlock's site file had 481 entries when we counted it on 24 September 2026. The project keeps an upstream list of sites excluded because they produce false positives, which the tool honours by default. --ignore-exclusions turns that off, and the help text says plainly that it may return more false positives. No API keys are needed.

Maigret: The Largest Database and Recursive Search

Maigret started as a Sherlock fork and has grown into the most capable of the group.

pip install maigret
maigret example_handle --html
maigret example_handle -a --timeout 15
maigret example_handle --tags gaming,coding

• Maigret needs Python 3.10 or later. Snap, Docker and a standalone Windows executable are also published.
• A default run checks the 500 highest-ranked sites by traffic from a database the README puts at about 5,900 sites (the data file held 6,121 entries when we counted on 24 September 2026). -a checks them all, which takes considerably longer and returns more noise from small sites.
• --tags narrows the run by site category or country code.
• --html (or --pdf, --json) writes a report. The Maigret documentation covers every option and the library interface.

Two things set Maigret apart. First, it parses the account pages it finds and extracts identifiers such as linked usernames and IDs, then searches for those too. This recursive search is powerful and noisy in equal measure: one wrong extraction sends the run down unrelated accounts, so read the chain before accepting it. Second, it fetches an updated site database from GitHub once every 24 hours, which means fixes reach you without a reinstall, and also that two runs a day apart can differ. Record the version and date with your results. Maigret's optional AI summary mode sends results to an OpenAI-compatible API and needs your own key; we would not put its output in a report without checking every line.

WhatsMyName: Data First, Web App Second

WhatsMyName is different. Since May 2023 the project has maintained only the detection data, wmn-data.json, and left the checking to other tools. On 24 September 2026 the file held 717 sites in 21 categories, each with a known-good username for testing. It is licensed CC BY-SA 4.0.

The quickest way to use it is the whatsmyname.app web tool mentioned above, which runs the checks from your browser, filters by category and exports CSV. For scripting, download wmn-data.json from the project's GitHub repository and read it directly:

jq -r '.sites[] | select(.cat=="finance") | .name + "  " + .uri_check' wmn-data.json

This lists the finance-category sites and their URL templates. It is also a good way to audit another tool: if a site you care about is in WhatsMyName but missing from your checker, you know where the gap is. WhatsMyName's inclusion rules explain its low false-positive rate: sites must be public, must include the username in the account URL, and must not transform the username. The trade-off is lower coverage than Maigret, and nothing on sites that sit behind logins or use numeric IDs.

Blackbird: A Fast WhatsMyName Checker

Blackbird exists and works, and it is worth being clear about what it is: a fast asynchronous checker built on WhatsMyName data, with username and email modes and PDF and CSV export. The repository has moved from its original account, and the old URL redirects.

git clone https://github.com/p1ngul1n0/blackbird
cd blackbird
pip install -r requirements.txt
python blackbird.py --username example_handle --csv

• git clone fetches the code (there is no PyPI package). The requirements file pins exact versions, so use a virtual environment.
• --username runs the checks and --csv or --pdf exports results. --no-update skips refreshing the site list at start-up.
• The optional AI feature needs a key created with --setup-ai. The README says only site names are sent.

Because it uses WhatsMyName data, Blackbird's coverage and accuracy follow that project's. Its last commit was on 13 July 2025, a dependency update. It works today, but check the date again before you build a process on it.

Social Analyzer: Scored Detection, Quiet Maintenance

Social Analyzer claims more than 1,000 sites and has a feature the others lack: a detection score from 0 to 100, which it groups into good, maybe and bad results.

pip3 install social-analyzer
python3 -m social-analyzer --username "example_handle" --filter good --output json

• --filter good keeps only high-scoring detections, which cuts noise considerably.
• --mode switches between fast, slow and special detection. Screenshots and the slower modes need a browser installed.
• --output json gives machine-readable results. A Node.js web app version is also available from the repository.

The friction here is maintenance. The GitHub repository's last commit was on 12 January 2026, but the PyPI package is still version 0.45 from May 2022. Site detections drift as sites change, so a four-year-old package will have broken rules. If you use it, install from the repository, and treat it as a second opinion rather than your main tool. It is licensed AGPL-3.0, which matters if you plan to build it into a product.

Sherlock vs Maigret vs WhatsMyName: How Do Coverage, Accuracy, Speed and Upkeep Compare?

All figures are as stated by each project or counted by us, and all dates were checked on GitHub and PyPI on 24 September 2026.

Speed. We do not quote timings, because they depend far more on your connection, your timeout settings and how many sites rate-limit you than on the tool. In practice, WhatsMyName-based checkers and Sherlock finish a default run quickest, Maigret's default top-500 run is comparable, and maigret -a or Social Analyzer's slower modes take much longer. Cloudflare challenges and rate limits cause more missing results than any tool difference, and they get worse if you run several tools back to back from one IP address.

Our recommendation. Use Maigret as the main engine when you need breadth and linked identifiers. Check its hits against WhatsMyName, either in the web app or through a WhatsMyName-based checker, when accuracy matters more than volume. Keep Sherlock for quick checks and for scripts where simplicity wins. Treat Social Analyzer as an occasional second opinion, and re-check Blackbird's maintenance before relying on it.

Where an Aggregated Platform Fits

Open-source checkers are excellent at one job. The friction starts around them: five virtual environments, different output formats, rules that change daily, results that live in text files on one laptop, and no record of which tool said what on which date. Then the handle leads to an email address, a phone number or a domain, and you need different tools again. Turning one handle into a map of related accounts takes more than one checker.

UserSearch 2.0 addresses those gaps. Username Intelligence checks a handle across platforms from the browser. OneScan runs one input across several selected sources and merges the results with source attribution, so you can see which source reported each account; the Credit cost is the sum of the selected sources and is shown before you run. Bulk search takes up to five handles through one Module in a batch, useful for spelling variants. In Forensic Mode, searches and bookmarks are kept in a Case, which gives your team the audit trail the command-line tools do not. From there, Email Intelligence, Phone Intelligence and Domain Intelligence are in the same account, with 100+ third-party data sources available without separate subscriptions. SargeBot can suggest next searches and draft a PDF report, which you verify. Results remain leads: coverage and freshness depend on each source, exactly as with the open-source tools. For a longer look at running a handle across thousands of sites, see our guide to mapping a username across 3,000 sites, and the username analysis knowledge base.

Combining Tools in Real Casework

Scenario One: Verifying a Spokesperson Account for a Newsroom

Context. A newsroom receives statements from an account, @example_handle, that says it speaks for Example Relief, a fictional charity. The editor wants to know whether the account is established or new.

Actions. The researcher runs Maigret's default top-500 check and WhatsMyName in the browser. Maigret reports 19 accounts, WhatsMyName 11. Nine appear in both. The researcher opens every hit. Of Maigret's ten extras, four are false positives (a site returning a generic page, two forum search pages, one login wall), three are genuine but inactive, and three are active and consistent with the charity's work. The two WhatsMyName-only hits are genuine: sites missing from Maigret's default top 500. The accounts' creation dates and linked website all point to example.org, the charity's real domain, and the oldest account dates back six years.

Outcome. The newsroom reports the account as long-standing and linked to the charity's own site, and cites the checks it made. The false positives are logged in the Case as "checked, not the account" so no one re-reports them later.

Scenario Two: A Marketplace Seller Handle in a Fraud Review

Context. The trust and safety team at Example Market Ltd is reviewing seller example_seller, reported by several buyers for undelivered goods.

Actions. Username Intelligence and OneScan return the same handle on two other marketplaces and a gaming forum, each with source attribution. Maigret's recursive mode extracts a second handle, example_seller_uk, from one bio, and the analyst runs that as a Bulk search alongside three spelling variants. Two of the other marketplace accounts carry buyer complaints similar to Example Market's. The gaming forum account turns out to belong to someone else with the same handle, as the language, dates and interests do not match.

Outcome. The team links three seller accounts with medium confidence, discounts the forum account, and shares the marketplace findings with the other platforms through their reporting channels. The Case holds the searches, bookmarks and the analyst's notes on why the forum account was excluded.

Auditing Your Checker Against Known Accounts

Before relying on any tool in a new case, run it against a handle you control or a known public account on the sites that matter to you. If it misses accounts you know exist, fix the gap before the real run. WhatsMyName's known field and Maigret's --self-check option exist for exactly this.

Handling Common Handles

Short or common handles, like alex or a first name plus a year, produce mostly unrelated accounts. For those, lean on the extras: consistent avatars, bios, linked sites, and creation dates that fit the timeline. Username matches are weak on their own and strong only when other details agree. Our advanced username OSINT guide covers bio and logo matching in more depth.

Using Username Checkers Lawfully and Proportionately

Account data about identifiable people is personal data, even when it is public. Before you run a batch, record your purpose, your lawful basis and why username research is necessary for it. For most commercial research the lawful basis is legitimate interests, which the ICO explains in its guidance on the legitimate interests basis, including the three-part test of purpose, necessity and balancing.

Keep it proportionate. Check the handles your question needs, keep only the results that matter, and delete the rest on schedule. Respect each site's terms and rate limits: the tools send many requests quickly, and pointing them at the same site repeatedly is poor practice. Do not log in to view restricted content, and do not contact account holders as part of research. Finally, report findings with confidence levels and keep the raw output so another analyst can check what you did. This is general information, not legal advice.

Frequently Asked Questions About Username Search Tools

What is the difference between Sherlock, Maigret and WhatsMyName?

Sherlock vs Maigret vs WhatsMyName is a choice between simplicity, breadth and precision. Sherlock is the quickest checker to install and script, with 481 site entries in our count. Maigret has the largest database, about 5,900 sites per its README, and extracts linked identifiers for recursive searches. WhatsMyName publishes careful detection data for 717 sites and runs as a web app.

Which username search tool is the most accurate?

WhatsMyName has the most careful detection data of the open-source username search tools. Each site entry sets the text expected when an account exists and when it does not, lists a known account for testing, and follows strict inclusion rules. The trade-off is lower coverage than Maigret, so many analysts check Maigret hits against WhatsMyName when accuracy matters.

Why do username checkers give false positives?

Username checkers give false positives when a site returns a normal page for any username, when a login wall looks like an account page, or when a detection rule relies only on the HTTP status code. Rules also break when sites change their pages. Every checker hit is a lead to open and verify, never a finding on its own.

How do you install Sherlock and Maigret?

Sherlock installs with pipx install sherlock-project and needs Python 3.9 or later, and Docker also works. Maigret installs with pip install maigret and needs Python 3.10 or later. Neither needs an API key for standard checks. Installing each tool in its own virtual environment or with pipx avoids the dependency conflicts these tools often have.

Is Social Analyzer still maintained?

Social Analyzer's GitHub repository had its last commit on 12 January 2026 when we checked on 24 September 2026, but its PyPI package is still version 0.45 from May 2022. Site detections drift as sites change, so the old package will have broken rules. Installing from the repository and treating Social Analyzer as a second opinion is the safer approach.

Is it legal to use username search tools in the UK?

Account data about identifiable people is personal data under UK GDPR, even when public. Teams using username search tools should record a purpose and lawful basis, usually legitimate interests with its three-part test of purpose, necessity and balancing. They should check only the handles the question needs, respect site terms and rate limits, and delete surplus results on schedule.

Choosing the Right Username Tool for Each Job

No single tool wins everything. Maigret gives you breadth and linked identifiers, WhatsMyName gives you the most careful detection data, Sherlock gives you a simple run you can script, and Blackbird and Social Analyzer have their niches with clear maintenance caveats. Whatever you run, open every hit, log the false positives, and record the tool version and date.

Stop guessing. Start researching with UserSearch. Run Username Intelligence and OneScan from one UserSearch account with access to 100+ third-party data sources, carry the handle into Email, Phone and Domain Intelligence without switching tools, and keep every search in a Case your team can review and turn into a report.