Why I built Wordy

I'm not a native English speaker, and I've been trying to memorize new words and learn their pronunciation for a long time. The pieces I needed were always scattered: a definition in one place, a translation in another, audio somewhere else, a way to drill what I'd learned in yet another tab.

Wordy is where all of that lives. You paste a batch of words you ran into today. You get clean cards back — definitions in your target language and your own, real example sentences, the synonyms you'll actually use, the antonyms that tell you where the word ends, and natural-voice audio for every word. You save them, drill them, and come back to them tomorrow.

Because it's just me, I can move quickly. If something doesn't work for the way I learn — or the way you learn — I can change it the same week. A new practice drill, a different voice, a fix for a definition that's slightly off: small enough team to ship, big enough vision to keep adding.

A few things I committed to before writing any code.

I work in three buckets, ordered by how confident I am they'll actually ship. Closer = more certain.

• SoonNative phrasing, CEFR levels, and a write-a-sentence drill — the next couple of weeks.
• NextSpaced repetition, more target languages, and the UI in your native language.
• Future A daily review email with a story built from your own words, and smarter history that surfaces what you struggle with most.

The full list, with the why behind each, lives on the roadmap.

Wordy was designed and built with Claude. Definitions come from Claude (Anthropic). Pronunciation comes from ElevenLabs. Everything else is hand-rolled Rails and Nuxt, and lives in two small repositories I deploy myself.

If a definition is wrong, if a feature you'd pay for is missing, if Wordy crashes — please tell me. I read every message, and every contact form submission goes directly to my inbox.