Effective: 7 September 2026
Picking a business name is harder than it should be. The name has to: mean something in your language, mean nothing offensive in the next three languages you'll expand into, be available as a .com or .io, be free on the social handles you actually use, and ideally mean nothing awful when a teenager in another timezone puts it in front of "fail." Most name generators solve one or two of these. Naminago tries to solve all of them at once.
Each language has a curated dictionary of ~2,000 to ~2,400 word fragments, hand-classified as adjective, noun, verb, or compound brand, and tagged with which industries and tones they fit. A pattern engine picks fragments based on your industry and tone, joins them with the right spacing for the tone (premium names get spaces; modern names stay CamelCase), and runs them through a quality filter that rejects slur-adjacent output, awkward consonant clusters, and known-trademark collisions. No LLM is involved in name generation. Every name is a deterministic combination of curated fragments.
At launch: English, Spanish, Portuguese, Hindi, Arabic, French, Russian, and Polish — each generating in its native script. Hindi is the first non-Latin-script language; Arabic is the first RTL; Russian is the first Cyrillic-script language; Polish is the first with Central-European diacritics. Traditional Chinese, Japanese, and Korean are next on the roadmap. Long-term target: 75+ languages. Each language lands with its own curated dictionary, not a machine translation of the English one.
When you ask us to check a name, we forward it to a workflow that queries the relevant registries: RDAP for .com and .io domains, public APIs for Instagram / X / TikTok / YouTube / App Store, and (coming soon) additional country-code TLDs. The checks are best-effort. We surface the result with a "this is a snapshot, double-check before you commit" disclaimer.
We don't sell or auction names. We don't charge for generated suggestions. We don't show one name in front of another for payment. We don't require an account. We don't train an LLM on your generator usage. The site is funded (when it is) by unobtrusive third-party advertising, clearly separated from generated content.
Naminago is a small project. The fragment dictionaries are curated by humans; the engine is a few hundred lines of JavaScript; the site is a static bundle. We're based in the open, not in a particular geography. For press or partnership inquiries, see the Contact page.
Near-term: Bengali, Indonesian, Urdu, German, Japanese, Marathi, Telugu, Turkish, Tamil, Vietnamese, Korean, Thai. Medium-term: 25 more languages across the remaining scripts. Long-term: full 75+ coverage, deeper availability checks, regional variants (es-MX, pt-PT) within the eight flagship languages. We publish what we're working on in the changelog and on the home page when something lands.