# Namedesk

> Domain research for the age of AI. Namedesk appraises domain names using signals built for the moment LLMs are doing the searching - not the old-school SEO-flipper heuristics. Core differentiators: an LLM attractor signal (whether models surface a name unprompted), historical-presence provenance, a proprietary AI-era scoring and pricing engine, and intent-routed verdicts (the same signals recommend opposite actions depending on whether you're building, monetizing, or investing).

Namedesk is for founder-buyers evaluating a name before they build on it, operators monetizing a type-in asset, and investors pricing portfolios. It reads domains; it does not broker them.

## Key concepts

- **LLM attractor signal** - whether the models surface a name on their own, unprompted. High emission means the name already lives in model memory.
- **Intent inversion** - the same signal flips meaning by intent. High LLM recall is a headwind for a new brand (you compete with model memory every time a user types the query) and a tailwind for a monetizer or investor (that same memory feeds type-in traffic).
- **Historical-presence provenance** - an offline read on a name's prior web presence, used for provenance without depending on live-DB comps.
- **Hallucination calibration** - a cross-model check that flags when the models confidently invent facts a name never had.

## Product

- [Landing page](https://namedesk.app) - overview, sample reports, pricing
- [Naming](https://namedesk.app/naming) - startup name generator: invent a shortlist for an idea, ranked by availability and AI-fit
- [API docs](https://namedesk.app/docs/api) - REST + JSON, same wallet as the dashboard
- [Sign in](https://namedesk.app/sign-in) - email-based auth, 6-digit code

## Sample research reports (publicly visible)

- [google.com](https://namedesk.app/research/google.com) - saturation-level recall, every model knows it
- [stripe.com](https://namedesk.app/research/stripe.com) - clean brand, high attractor rate
- [notion.so](https://namedesk.app/research/notion.so) - dictionary word on alt-TLD, brand-vs-concept confusion
- [linear.app](https://namedesk.app/research/linear.app) - adjective on product-TLD, intent-routed verdicts side-by-side

## Editorial

- [The appraisers price the string. Buyers pay for what the name already is.](https://namedesk.app/insights/appraisers-price-the-string) - ten expired-domain sales vs the Dynadot appraisal vs our blind band: 8 of 10 inside our band, Dynadot within 2x of the price twice
- [We invented 50 startup names. AI thought 38% of them were tech.](https://namedesk.app/insights/ai-confidently-wrong) - what frontier models do when asked about names that don't exist
- [The 8% of startup names AI hasn't decided about yet](https://namedesk.app/insights/blank-canvas) - the disagreement-tail names that builders should learn from
- [Name coining methodology](https://namedesk.app/insights/name-coining) - how Namedesk invents names that survive

## TLD Radar

The public record of the ICANN 2026 new-gTLD round - every company and community applying to run a new top-level domain, with a source-linked history and an AI read (a per-TLD score, 0-100) on which endings matter. Free, no auth.

- [Stats](https://namedesk.app/stats) - domain industry stats: every TLD and registrar month by month, from the registries' monthly reports to ICANN and from country-code registries that publish their numbers, with per-TLD pages at /stats/tlds/{tld} and per-registrar pages at /stats/registrars/{iana-id}.
- [TLD Radar](https://namedesk.app/tlds) - the scoreboard: search and sort every applied-for TLD string, each linking to its dossier.
- [Dossier pages](https://namedesk.app/tlds) - one page per ending at /tlds/{ending}, with who's applying, the evidence, the full status history, and the AI score.
- [Most contested TLDs](https://namedesk.app/tlds/contested) - the ranked cut of the record: every applied-for TLD string two or more applicants are chasing, most contested first.
- [Beat the AI](https://namedesk.app/tlds/play) - a free guessing game: score a new TLD from 1 to 100, then see the AI's score for it, one ending after another.
- [Radar changelog](https://namedesk.app/tlds/changelog) - the append-only feed of every application status change on the record.
- [Methodology](https://namedesk.app/tlds/methodology) - how the per-TLD AI scores are produced and what they mean.

Feeds (machine-readable, no auth):

- RSS: https://api.namedesk.app/api/tld-radar/changelog.rss - the changelog as an RSS 2.0 feed.
- CSV: https://api.namedesk.app/api/tld-radar/applications.csv - the full application table as CSV.

## Reference

- [FAQ](https://namedesk.app/faqs) - what Namedesk does, pricing, accuracy, privacy of research, submission size, rate limits, supported TLDs, report freshness, how to get help
- [Privacy](https://namedesk.app/privacy) - what we collect, how we use it, the rights you have over it
- [Changelog](https://namedesk.app/changelog) - every meaningful change ships alongside a changelog entry

## Pricing

Credit-pack model. One research report consumes 10 credits. Reports are cached at the domain level - researching a previously-researched name re-serves the cached report without spending credits.

- Starter - 100 credits for $5 (50c/report)
- Regular - 500 credits for $20 (40c/report)
- Desk - 1,000 credits for $30 (30c/report)

New accounts receive 10 free credits on signup - enough for one full research.