Dataset · 3,000 names checked in 8 extensions

All 1,000 of the most common English words are taken as .com domains

On 5 and 6 October 2026 we checked 1,000 common English words, 1,000 random two-word names and 1,000 invented words in .com, .net, .org, .co, .io, .ai, .app and .dev: 24,000 checks against the registries’ own servers. Every common word was registered in .com, .net and .org; only 39.0% of the two-word names were in .com. The per-name results are free to download and reuse with credit. The same words in 71 extensions are a separate study.

Download the CSV

3,000 rows · 10 columns · 381 KB · CC BY 4.0

Archived on Zenodo with the scripts that produced it: doi:10.5281/zenodo.23288857

Share of names already registered

1,000 names per sample. For names without nameservers, these shares are corrected with the RDAP checks, as published in the guide.

Share of names already registered, by extension and sample, October 2026
ExtensionCommon wordsTwo-word namesInvented words
.com100%39.0%58.5%
.net100%7.6%15.9%
.org100%6.9%12.6%
.co96.0%3.0%7.3%
.io98.4%1.6%3.9%
.ai99.9%1.3%4.0%
.app92.5%1.8%4.4%
.dev77.3%0.7%3.2%
  • Single words are gone everywhere. All 1,000 common words are registered in .com, .net and .org, and 70.3% of them in all eight extensions.
  • Combining words reopens the namespace. 39.0% of the two-word names are registered in .com, and no other extension passes 7.6%.
  • A free match almost always exists. 99.7% of the taken two-word .coms, and 99.5% of the taken invented ones, were free in at least one of the other seven extensions.

What is in the file

name
The name without its extension, for example primecotton.
sample
common_word: the 1,000 most frequent English words of three to ten letters that the WordNet dictionary lists, without proper nouns. two_word: 1,000 random pairs of a common adjective and a common noun. invented: 1,000 random pronounceable strings of five or six letters that are not English words. The random samples were built with fixed seeds.
com, net, org, co, io, ai, app, dev
What we saw for the name in that extension:
  • registeredThe registry delegates the name to nameservers, or its RDAP record shows it registered (61 of the registered results had no nameservers).
  • not_registeredNo nameservers, and the registry's RDAP database confirms the name is not registered.
  • no_nameserversNo nameservers, not checked in RDAP. Most such names are free, but not all: in our RDAP samples 0% to 2.4% of two-word and invented names without nameservers were registered, and up to 10 of 26 common words in .io.
  • unknownBoth checks failed (10 of the 24,000 results).

How it was collected

Each name was looked up in each extension with a non-recursive query to that extension’s authoritative nameservers: a delegated name is registered. Because a registered name can lack nameservers, every .com without nameservers was confirmed in the registry’s RDAP database, as were random samples of up to 60 such names per sample in each other extension. One registry limited our RDAP requests, so the invented-word results for .io and .ai rest on the DNS check alone.

Limits

This is a snapshot of two days. Registrations change every day, so a row describes the namespace in October 2026, not the status of a name today. “Registered” includes names a registry holds back for itself. The file keeps each result as observed, so shares computed from it differ slightly from the corrected shares above. For any single name, the registry’s own RDAP record is the final word.

Reuse and credit

The data is published under Creative Commons Attribution 4.0: you may share and adapt it for any purpose, with credit and a link to this page. Suggested citation:

Findadomain.name (2026). Taken domain names: 3,000 names checked in 8 extensions (October 2026) (Version 1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.23288857

The Zenodo record also holds the estimated shares as a CSV, a README with the method, and the Python scripts that built the samples, ran the checks and produced both files.