Show HN: Zenòdot – Find if a book has been translated into your language
Posted by AusiasTsel 1 day ago
I'm a multilingual reader (Catalan/Spanish/English/Italian), and I kept hitting the same wall: I'd hear about a book and have no way to know if it existed in my language. Turns out this is a genuinely unsolved problem. There's no single database that tracks which books have been translated into which languages. ISBN registries are fragmented by country. Open Library has great English coverage but gaps elsewhere. Wikidata has surprisingly rich translation data but it's locked behind SPARQL. Google Books is inconsistent across regions.
So I built Zenòdot to cross all four and piece the picture together.
What I found building it:
-The ISBN system is far more broken than I expected. ISBNdb has millions of English records but almost nothing for languages like Basque, Icelandic, or Bengali. Books exist in these languages, they just don't exist in the databases.
-Wikidata was the biggest surprise. It has structured translation data for thousands of works, but extracting it requires SPARQL queries, title resolution across scripts (try matching a book title in Chinese to its English original), and author alias caching. Hard to build, but the results fill gaps that no other source covers.
-The most interesting output isn't what the tool finds; it's what it doesn't find. When someone searches for a book in a language and there's no result, that's a demand signal. "Someone in the world wanted this translation and it doesn't exist." That data could be genuinely useful to publishers.
The tool prioritizes your selected languages, so it shows you editions relevant to you first. The philosophy is "documentary infrastructure”: no recommendations, no social features, no accounts. You search, you find (or don't), you go buy the book wherever you want.
Stack: Next.js 15 (App Router), Supabase, Vercel, TypeScript. Solo project, no funding, about 4 months of work.
If you're multilingual or learning a language, I'd especially love your feedback. Try searching for a book you love and switching between languages, that's where the tool shows its value.
Comments
Comment by mosbyllc 13 hours ago
Comment by AusiasTsel 8 hours ago
Comment by zufallsheld 19 hours ago
Comment by AusiasTsel 19 hours ago
Anna's Archive ISBN visualization is fascinating; it really shows how fragmented and incomplete the ISBN landscape is. We don't use their data directly (licensing concerns), but it confirmed what we were seeing: massive gaps in non-English coverage.
WorldCat is great for library holdings but harder to use for translation discovery specifically, it tells you "this library has this edition" but doesn't easily answer "has this book been translated into Basque?" across its whole catalog. We'd love to integrate it eventually, though.
Right now Wikidata turned out to be the secret weapon: it has structured translation relationships that none of the others provide.
Comment by MonkeyIsNull 23 hours ago
However, it stills looks like it's tied to localhost? I get a web socket error (to localhost:8081) and when I type the name of the title into the search title. Nothing happens and the button doesn't look like I can click it (no highlighting)
Comment by AusiasTsel 23 hours ago
Comment by MonkeyIsNull 23 hours ago
That's from an incognito window on Chrome
Comment by AusiasTsel 23 hours ago
Comment by MonkeyIsNull 23 hours ago
Yeah, I've tried switching the languages, not ALL the text on the page changes from the default language and I've tried not changing the lang. Same thing.
Comment by AusiasTsel 22 hours ago
Comment by AusiasTsel 22 hours ago