Checked July 3 to 5, 2026: about 750 names
Every real dictionary word, obvious two-word combination, familiar phrase, and Latin single word was already registered or parked. Asking prices ran from $20,000 to $39 million.
The free .com pool is effectively gone. Here is what that means for finding a name, with the date each number was checked.
Nearly every good real-word .com is already taken. More than 3,000 names were checked in early July 2026, and every real word or natural combination was registered. The names still free do not read as real brands. So you buy a good .com, and it costs less than the early estimates claimed.
Every real dictionary word, obvious two-word combination, familiar phrase, and Latin single word was already registered or parked. Asking prices ran from $20,000 to $39 million.
Single words, obvious combinations, hand-crafted names, and more than 1,000 built pairs. Every name that felt right was taken. The only names still free did not read as real brands.
The earlier advice to mine weaker word patterns produced mediocre names that failed with buyers. Do not chase those for a great name.
One lane is left: combine two real words from unrelated fields. Otherwise you buy a two-word brandable, typically $1,000 to $4,000, or catch an expiring .com at auction. DropCatch catches 60 to 80 percent of moderate names for about $59 to $70, and running NameJet alongside raises the combined catch by about 35 percent.
The earlier average of about $23,000 was inflated. A dataset of 480,000 sales puts good brandables in the low four figures. Cheap does not mean mediocre; judge the name itself, not the price.
Open opportunities close within days once they are discoverable. Two names were registered days before the search found their lane. If a name shows as available, register it the same day or lose it.
Any list of available names is a snapshot. Re-check the whois registration record before you act.
These numbers come from the doctrine, which explains the method and how it dates its own evidence.
Doctrine
DoctrineThe naming doctrine, published verbatimThe processSeven steps from a buyer to a registered domainThe gatesTen hard laws and seventeen mechanical checksRegisters & tasteThe three registers and the taste corpusGeneration methodsHow candidates are producedTLDs & the aftermarketThe ranked TLD policy and the aftermarketLegal & post-lockThe knockout legal gate and post-lock governanceAnti-patternsThe anti-relapse guardrailsToolchainRegistry truth, batch patterns, generatorsSOTA methodologyThe 2026-09-16 research sweepResearch sweep, Sept 2026The 2026-09-30 delta against the doctrineSearch the doctrineFull-text search across the doctrineTools
Single appraisalScore one name, twelve checksBulk analysisScore a list of names at onceName generatorGenerate names against the doctrineCold-read probeWhat a zero-context model guessesMarket
DroppingWhat is about to drop, ranked by soundlikenessTLD economicsRenewal-first registrar costs and the promo trapMarket dataSales, pricing, TLDsMarket trendsWhat is movingReportsPeriodic written analysisGuidesMethod, worked end to endReference
AboutAbout the projectPricingPlans and pricingBlogWriting and updatesHelpHelp and documentationFAQFrequently asked questionsAPI accessAPI keys and accessSupportGet supportContactContact usTeamThe teamCareersOpen rolesWebinarsUpcoming and recorded webinarsLegalLegal overviewPrivacyPrivacy policyTermsTerms of serviceCookiesCookie policyNo results for this search.