FAQ
The questions buyers ask before signing up, the answers we give them.
What does Namedesk actually do?
Namedesk reads a domain name and tells you what leading AI models think about it. Each report bundles dozens of distinct signals, grouped into three focal areas.
The AI dimension — recall across several frontier AI models, unprompted emission (whether models surface the name on their own), cross-model hallucination calibration, and industry and intent inference.
Market signals — a tiered pricing band with reasoning, key factors, and risks; spellability (whether an AI agent reaches your name when a buyer describes what they want); and lookalike collision detection across typo neighbours and brand confusables.
Brand and trust — brandability with nine sub-evaluations (recall, confusion, brand, voice, pronounce, typeable, memorability, fit, and trademark risk), commercial-intent markers, tokenization, and the deterministic name-shape features that anchor everything: character length, TLD and TLD tier, pattern class, syllable count, dictionary likeness, geographic markers, and short-premium status.
The verdict at the top of every report is written for the intent you pick — building a brand, monetizing traffic, investing in a portfolio, or exploring options. Same signals, four framings.
Who is this for?
Founders naming a company who want to know which names already have a crowded place in AI memory before committing to a brand.
Investors valuing a domain or portfolio who want a signal that isn't on every comparison sheet yet — an AI-era read on the name, not just a blind average of stale registrar comps.
Anyone who doesn't have a name yet. Use the coining feature to invent a shortlist for your idea — paired with the TLDs that fit, ranked by what's still available.
Namedesk is not built for bulk flippers, registrar farms, or pay-per-thousand lookups. There are cheaper tools for that.
How does the pricing work?
Pay as you go. Credit packs run from $5 (100 credits, 5¢ each) to $30 (1,000 credits, 3¢ each). No subscription, no expiry. New accounts receive 10 free credits on signup — enough for one full research, one coining run, or one round of alternatives.
Three things spend credits:
Researching a name — 10 credits reads one domain. Reports are cached at the domain level, so if a name has been researched before you read the cached report and don't spend credits.
Coining a shortlist — 10 credits invents a batch of names for your idea, paired with TLDs and ranked by availability.
Drafting alternatives — 10 credits, on top of an already-researched name, generates a fresh slate of alternative names in the same space.
Do you offer volume or enterprise pricing?
Day to day it's simple pay-as-you-go — the credit packs on the home page, where the bigger the pack, the cheaper the read. Every feature, the dashboard and the API alike, is the same product at any scale; there's no tier gate to unlock.
If you're running names at real volume — a large portfolio, an agency workflow, or heavy API traffic — email hello@namedesk.app and we'll put together custom or enterprise pricing for your usage.
How accurate is the appraisal?
Every read draws on multiple frontier AI models. A single model's answer is an anecdote; agreement across models is a signal. The pricing band comes from our own engine, which weighs the AI-era read against real comps — not a blind average of stale registrar prices.
We tell you when confidence is low. If signals conflict or the name has no nearby anchor, the report says so plainly. We'd rather be honest about uncertainty than confident and wrong.
Does Namedesk flag negative or off-putting associations?
It can. Because the read comes from frontier AI models rather than a fixed word list, it picks up nuance: if a name sounds like something unfortunate, leans on a controversial meaning, or sits too close to an association you wouldn't want next to your company, that tends to surface in the AI and brand signals, and the verdict takes it into account.
We don't keep a blocklist of "bad" words — the models weigh meaning and context the way a person would, which is what catches the subtle cases a keyword filter misses. Like every signal, treat it as a strong prompt to look closer, not a guarantee.
What if an AI model is having an outage?
Reads run with built-in tolerance — the report still goes through when a majority of the model providers respond in time. If too many providers are down at once we hold the read and try again, rather than shipping a half-formed report.
Is my research private?
Your account history — which domains you've read — is private to you. The report content for a given domain is shared across the platform: if you research a name and someone else asks about the same name later, they read the cached report. We don't tell anyone that you were the one who triggered the read.
When we send a domain to a model provider for analysis, we send only the domain string. Your identity, email, and account details never leave Namedesk.
How long does a report take?
A solo read finishes in a couple of minutes. The dashboard streams signals in zone by zone so you see the report take shape rather than waiting at a blank page. Bulk submissions of up to 100 domains process concurrently and typically finish within a quarter of an hour.
Can I share a report?
Yes. Every report has a public share link. The share view shows a curated subset of the report and a clear path for the recipient to research their own names — useful for sending a verdict to a co-founder or an investor without giving them dashboard access.
Do you have an API?
Yes. Same wallet as the dashboard, REST + JSON, public documentation at /docs/api. Generate a key in Settings → API keys.
Can I export a report?
Yes. Every research report has a one-click Markdown download in the dashboard — structured so you can paste it straight into ChatGPT or Claude as context for a follow-up. Through the API, every report comes back as JSON in the same shape as the dashboard, and bulk pricing results export as CSV for a spreadsheet. Whatever you see on screen, you can take with you — by hand or programmatically.
We don't currently offer white-labeled or co-branded reports for client presentations. If that's a blocker for your work, write to us — it's the kind of thing we'll prioritize when buyers ask for it.
How big a submission can I make?
Up to 100 domains per submission, whether through the dashboard or the API. Larger jobs need to be split. We don't auto-chunk because we want each submission to be a unit you can track and reconcile against your wallet.
If 100 domains is the wrong cap for your workload, write to us before working around it. The cap is for fairness during traffic spikes, not because we don't want your traffic.
Are there rate limits?
For API requests, no per-key request ceiling in normal operation — your spend is bounded by the credits in your wallet. We do throttle very high-concurrency bursts on a single account so one customer's backlog never starves another's traffic.
For sign-in, a small number of attempts per minute from the same network address. Reach the limit and sign-in pauses for the rest of the minute, then resets. Designed to deter brute-force guessing without affecting normal use.
What kinds of domains do you accept?
Any domain with a valid TLD. The name doesn't need to resolve, be registered, or be available — appraisal works on any name shape.
We don't currently accept internationalized (Punycode) domains. If you need them, write to us — it's a roadmap item, not a permanent restriction.
How fresh are the reports?
Cached reports serve immediately and don't consume a credit. When we ship a meaningfully better version of the underlying analysis, the next read on that domain refreshes the report (and consumes a credit). You can also force a refresh from the API.
Are reports guaranteed? Do you offer refunds?
Reports are best-effort. The signals we surface come from external AI models whose outputs change over time, and Namedesk does not warrant that any specific result will be accurate, complete, or definitive. Treat a report as one strong input to your decision, not the decision itself.
All credit purchases are non-refundable. Credits already spent on a report, a coining run, or an alternative-names run are not returned. If something is clearly broken — a failed run that didn't deliver, a duplicate charge, an obvious bug — write to us and we'll make it right at our discretion.
Do you offer an SLA or uptime guarantee?
No. Namedesk is pay-as-you-go and best-effort — no SLAs, no guaranteed response times, and no priority-support tiers. Reads are built to absorb a model provider being down: a read still completes when most providers respond, and holds rather than ship a half-formed report if too many are out.
Meaningful changes go on the changelog, and hello@namedesk.app reaches us within a business day, usually sooner. If you need contractual uptime or support guarantees, we're not the right fit today — but tell us what you'd need.
How do I get help?
Email hello@namedesk.app. We respond within a business day, often within a few hours. Material product changes are announced on the changelog.