NameHound

Privacy

Short version: no accounts, no cookies, no advertising or tracking scripts, cookieless visit counts from Cloudflare Web Analytics, and no log of the names you check. The only thing you can send us is a suggestion or a wrong-result report, and only if you choose to. One button on the band name generator sends a name from your browser straight to Wikidata, and only when you tap it. A free read-only connector for AI assistants checks names the same way and keeps only daily totals (below). Here is exactly what happens.

Last updated 2026-10-09. We will update this page before we change any of it.

When you check a name

  1. Your browser sends the name to our server at namehound.app, once per automatic check.
  2. Our server passes the name, and nothing else about you, to the registry or platform being checked: domain registries (through RDAP), Bluesky, three Mastodon servers, GitHub, GitLab, npm and PyPI, and, once switched on, YouTube, Twitch and Steam. Each of those services handles the request under its own privacy policy. They see our server's address, not yours.
  3. The answer is stored in a short-term cache for up to 10 minutes so a repeat search doesn't ask the platform again. Then it expires. The cache holds the name and the result only.

We don't write searched names to logs or a database (unless you type one into the suggestion form yourself), and we've turned off request logging for the server. The click-to-check links send nothing through us: your browser goes straight to that platform, and that platform's privacy policy applies.

AI assistant tools

NameHound offers name availability checks (domains and handles) and our Brand Name Score to AI assistants such as ChatGPT and Claude, through a connector at https://namehound.app/mcp. When an assistant uses it, the assistant's servers send us the name to check. We use that name only to answer that one request. The only copy we keep is the short-term cache described above (up to 10 minutes, name and result only), which these checks share with the website; after that it is gone. We don't log it, write it to a database, or link it to you. As on the website, our server passes the name, and nothing else about you, to the registries and platforms it checks (domain registries through RDAP, Bluesky, three Mastodon servers, GitLab, npm and PyPI). The assistant app you use sends us the name and receives our answer. The assistant connector has its own small daily limit for those lookups, so it can never use up the checks the website needs. The connector has no accounts and no keys, and it does not check GitHub.

What we keep: daily totals only. For each day we count how many times the tool was used, which kind of assistant app it came from (ChatGPT, Claude or other), whether it worked, a rough count of results (none, one, two or three, or four or five), and how many visits came to our site from links in the answers (those links end in ?ref=agent and the tool's name). We also keep a few housekeeping totals per day: how many requests our limits turned away, how many lookups the connector made to registries and platforms against its daily limit, and how often a few particular answers were served (to spot copying). The totals contain no questions, names, IP addresses or browser details, so they can't be traced to a person. We keep the daily totals for as long as the site runs. If the site is ever sold, they pass to the new owner.

How we count visits from those links: a link in an answer opens this site with ?ref=agent and the tool's name in the address, and the name you checked after a # (which your browser never sends to a server). When our server sees that tag on a page request, the daily total for that tool goes up by one. Nothing else is read or stored about you.

Who else handles it: our host, Cloudflare, runs the connector and processes each request to deliver it. To stop abuse, Cloudflare's rate limiter counts requests per IP address for about a minute; we don't store that count. Cloudflare's request logging feature (Workers Logs) is turned off for this site, so no per-request log of connector calls is kept through it. The assistant app you use (for example ChatGPT or Claude) has its own privacy policy for what you type into it. We don't sell or share anything from these requests, and the tool shows no ads.

Extra details an assistant may attach: an assistant can attach technical hints to each tool call, such as your language setting, an app version hint and a coarse location. The tool does not use or keep any of these (it reads only the app's name, to count ChatGPT, Claude or other).

Your choices: there are no accounts, and nothing we keep can be linked to you, so there is nothing to look up or delete. You can stop using the connector at any time in your assistant's settings. Questions: support@namehound.app.

Band name generator: Ask Wikidata

On the band name generator, the check panel has an Ask Wikidata button. It does nothing until you tap it. When you do, your browser (not our server) sends the name, plus a few spellings of it (with and without "The", and with "&" or "and"), to the Wikidata Query Service at query.wikidata.org to ask whether a band or musician already has that name. The request goes without cookies and without telling Wikidata which page you came from. It does say the request comes from NameHound: your browser adds our site's address, and we label the request NameHound/1.0 as Wikimedia asks. It does not pass through us, so we don't see or keep it.

Because your browser makes the request, Wikimedia receives your IP address and standard browser details, as with any site you visit. Its privacy policy says it covers Wikimedia "Application Programming Interfaces (APIs)" and that it automatically receives "the IP address of the device (or your proxy server)" you use. If you'd rather not share that, skip the button and use the MusicBrainz and Bandcamp search links, which you open yourself.

Your IP address

To stop abuse, our server counts how many checks each IP address makes per minute. The count is kept by Cloudflare's rate limiter for about a minute and isn't stored by us. Our host, Cloudflare, also processes standard connection data to deliver and protect the site; see Cloudflare's privacy policy.

Visit counts

We use Cloudflare Web Analytics to count visits and to see which pages are used and how fast they load. It does not see the names you check, which stay after a # in the address. Cloudflare says its Web Analytics does not use any client-side state, such as cookies or localStorage, to collect usage metrics, and that it does not fingerprint individuals via their IP address, User Agent string or any other data for the purpose of displaying analytics. It records page views, the site that referred you, your country, browser, operating system and device type, and page load performance. Cloudflare processes this under its privacy policy.

In your browser

The name you typed is put in the page address after a # so you can bookmark or share a search. Browsers don't send that part of an address to any server. Your trademark-search answer stays in the page and is gone when you close it.

YouTube

When the YouTube check is switched on, it uses the YouTube API Services. By using that check you agree to be bound by the YouTube Terms of Service, and Google's handling of the request is covered by the Google Privacy Policy. We store YouTube results only in the 10-minute cache described above.

Suggestions and wrong-result reports

The suggest a platform or report a wrong result page lets you send us a message. When you press send, we store only:

Why: so we can re-check a result against the platform's official interface, or see whether a platform offers an official way to check names. We delete the message itself 12 months after you send it. These records are stored in a Cloudflare D1 database (Cloudflare is our web host) and processed under Cloudflare's privacy policy. There is no CAPTCHA and no third-party spam service: a hidden field that only bots fill in and the rate limit do that job.

To sort messages we use a small AI model, run through OpenRouter, to label them by type. Only the message, the platform, the link and the page it came from are sent for that; never your email or the IP code.

To have a message (and any email you left) deleted sooner, send another message through the same form with the type "Something else", saying roughly when you sent the first one and what it was about. We delete both. We can't look a message up by your name, because we don't ask for it.

What we don't do

Contact

Questions about privacy: hello@namehound.app.