Search Newly Registered Domains Without Downloading a Single List

By DomainKits Team • July 4, 2026

Here is how the newly registered domains (NRD) industry has traditionally worked: every day, providers publish a compressed list of yesterday's registrations. If you want to know anything, you download the file, unzip it, and search it yourself. Want last week too? Download seven files. Want a 60-day view? Sixty files, at 200,000+ domains per day, and now you are maintaining a 12-million-row dataset to answer one question.

And the question was probably something like this:

• "Did anyone register a domain containing my brand this week?"
• "Are people registering names around this keyword, and is it accelerating?"
• "What did this registrant pattern look like over the past two months?"

These are search-shaped questions. The industry keeps answering them with download-shaped products, built for data engineers with pipelines, not for the brand owner, investor, or analyst who just needs an answer before lunch. DomainKits Search New exists to close that gap: the past 60 days of registrations across 1,000+ gTLDs, held in one searchable index, updated daily.

What you can ask it

A search box alone does not solve the problem; the filters are what turn a domain list into answers. Each filter exists because of a question people actually bring to NRD data:

• Keyword, with position control. Search names containing your term, and specify whether it appears at the start, middle, or end of the name. You can also exclude noise terms (test, demo, tmp) in the same query.
• Days since registration (0-10, 10-20, 20+). Freshness bands. The newest band is where monitoring work lives; the older bands show whether a pattern is persisting or fading.
• TLD, plus cross-TLD registration count. Filter to one extension, or read the result the other way: each domain shows how many TLDs the same name was registered in. A name claimed across five extensions at once is a deliberate act, and that is visible at a glance.
• For-sale status. Names already listed for resale identify speculative holdings, whether you are hunting acquisition targets or measuring how much of a trend is inventory rather than intent to build.
• Character type and length. Pure alphabetic, no numbers, no hyphens, under 5 characters, and similar cuts. These are the quality filters investors apply mentally anyway; here they are checkboxes.
• Registration period. How long the registrant paid for. A multi-year registration signals a different level of commitment than a minimum one-year term.

What each result tells you

Every domain in the results carries its own mini-profile, so different readers can take different things from the same list:

• The name, parsed into words. Each domain is shown with its segmented reading: cloudnestrobotics.com appears as "cloud nest robotics" (powered by DKSplit, our open-source model trained on EuroHPC's Leonardo supercomputer). For anyone scanning hundreds of results, this is the difference between reading names and decoding them.
• Freshness and creation date. Domains registered today are flagged as such, alongside the exact creation date.
• Registration period. How many years the registrant paid for up front.
• Cross-TLD registration count. How many other extensions the same name exists in. A name claimed across several TLDs at once was claimed deliberately.
• One-click follow-ups. WHOIS and related lookups per result, and a Make Offer path for names you want to pursue.

Registered users can export any filtered view.

Three searches, three jobs

The brand check. Search your brand as a keyword, no other filters, sorted by freshness. This is the weekly two-minute routine that replaces "we found out when the phishing email went out." Anything suspicious gets the follow-up treatment: WHOIS, DNS, and safety checks are one click away in the results.

The trend check. Search the keyword you suspect is heating up, then read the shape of the results: how many names, in which TLDs, how fresh, how many already for sale. You are not just counting registrations; you are reading their composition.

The prospecting pass. Combine a keyword with quality filters: pure alpha, under 10 characters, not hyphenated, for-sale status off. What remains is the set of clean, recently claimed names in your niche, which is the raw material for both acquisition lists and naming research.

One search, three interfaces

The search box is the human interface, but it is not the only one. The same index is reachable at three altitudes, depending on who, or what, is asking:

• The search page, for a person who needs an answer now. Everything described above, no setup.
• The REST API, for code. GET /api/v1/search/nrds exposes every filter on this page as a query parameter: keyword with position, days_range, period, has_sale, no_hyphen, and the rest, so a monitoring script or dashboard can run the exact search a person would. Available on Premium and higher plans, and the same API key drives the DomainKits MCP Server, which means an AI agent can run these searches as a native tool.
• The daily download, for pipelines. If you are running your own classification, enriching a security stack, or joining NRD data against internal datasets, you want the raw stream, and that is what the daily NRD download is for.

Search when a person needs an answer, API when a system asks the same question on a schedule, download when you need the whole stream.

But if you have ever unzipped a domain list, opened it in a spreadsheet, watched the spreadsheet struggle, and thought "I just wanted to check one name", the search box was built for you. Try it with your brand name; that first search is the habit worth forming.

Related reading: for what registration patterns can reveal about the people behind them, see Reading Registration Intent in Newly Registered Domains.