MyRecon vs Sherlock

What Sherlock is

Sherlock is an open-source Python program, MIT licensed, that takes a username and asks a long list of websites whether a profile with that name exists. Its README describes coverage of more than 400 social networks. You install it with pipx, Docker, or a package from Kali, BlackArch, Homebrew and several Linux distributions, then run it from a terminal. Hits print as a list of URLs and can be written to text, CSV, XLSX or JSON. It supports a proxy, a configurable timeout, an opt-in flag for adult sites, and a mode that opens every hit in your browser.

For a lot of people Sherlock is where username OSINT starts, and it earned that. The site list is maintained in the open, anyone can add a site with a pull request, and it runs anywhere Python runs.

Where Sherlock beats MyRecon

• Reach. 400+ sites against our 123 on the web. If a handle lives on a niche forum, Sherlock has a better chance of knowing that forum exists.
• Scriptability. It is a command you can put in a loop, a cron job or a pipeline. Our web tool is a page; our own CLI exists, but Sherlock's has years of community tooling around it.
• Network control. A proxy flag means you choose where the requests come from. MyRecon's web requests come from our server, which some sites block outright.
• No ads, no rate limit. It runs on your machine, so the only limit is how politely you choose to query.

Where MyRecon beats Sherlock

• Nothing to install. Open a tab, type a handle. That matters more than it sounds: most people who need a username check once will never set up Python to do it.
• A confidence label on every match. Sherlock's per-site rule is a yes or no. MyRecon scores status, page content, profile metadata and redirects together and labels each match high, medium or low, so you know which ones to open first.
• It explains its rejects. Under every result is a panel listing the platforms that answered but were not reported, with the status code and the reason. Sherlock can print its misses with --print-all, but a blocked site and a missing account look the same there. We print the difference.
• One page for the whole case. A handle leads to an email, the email to a domain, the domain to an IP. MyRecon runs all of those lookups on the same page and turns each finding into a one-click pivot. Sherlock does one job and leaves the rest to you.

The false-positive question

Every username checker is guessing, in a disciplined way. A site that answers "200 OK" for any name you type, real or not, will produce a hit for every search. Sites change their pages, add login walls and start blocking datacentres, and a rule that was correct last month quietly starts lying. Sherlock's maintainers fix these as reports come in, and its issue tracker is full of exactly this work. That is not a knock on Sherlock; it is the nature of the job.

We measured the scale of the problem ourselves. When we tested 1,015 candidate sites from public lists against a real handle and an invented one, 66 answered 200 to both, which makes status-code-only detection a permanent false-positive source for those sites. The method is written up in our guide on why username checkers report fake accounts. Our Android app goes a step further and requests an impossible control handle alongside the real one on every platform, so a site that answers both the same way cannot produce a match at all.

Whichever tool you use, run the control test yourself: search a handle that cannot exist. Every result it returns is a false positive, and it tells you how much to trust the rest.

Privacy and where the requests come from

Sherlock sends its requests from your own IP address unless you give it a proxy. That is good for control and bad if you do not want the target's platforms seeing your address. MyRecon's web search sends them from our server, so the sites see us, not you. We do not store searches in a database; history lives in your browser. The Android app is the reverse again: requests go from your handset, which is why it reaches more platforms, since phones are blocked far less often than servers.

Which one should you use?

If you are a researcher who already has a terminal open, install Sherlock, and use MyRecon when you want a second opinion with confidence labels, or when you need to move from the handle to an email or a domain without switching tools. If you are checking your own footprint, or verifying one person once, MyRecon gets you there with less friction. Both are free and both are open source, so the cost of trying the other is a few minutes.

Questions people ask

Other comparisons