crosswort

Daily crosswords where the clues are in a foreign language

Visit Website
August 11, 2026 After 4 years of doing NYT crosswords I built a crossword app to trick myself into learning Spanish.

I'm terrible at learning languages. Not for lack of trying — I've started and abandoned more apps than I can count. And I could never figure out why the vocabulary just wouldn't stick, until I noticed the pattern: every app I used tested recognition. Four options, tap the right one. I'd feel great, my streak would climb, and then in an actual conversation I couldn't produce the word to save my life.

Recognition isn't recall. Being able to pick a word out of a lineup is a completely different skill from pulling it out of your head cold when you need it. And almost every popular app optimizes for the first one because it feels like progress and keeps you tapping.

The one format that always forced real recall for me was crosswords. You get a clue and a few crossing letters, and you have to retrieve the answer from nothing. So I built the thing I actually wanted: a crossword app where the clues are written in the language you're learning, and the answers you fill in are in English. Solving a puzzle means reading and decoding the target language, then producing the meaning — never picking it off a menu.

That's Crosswort. Where it is now:

  • 12 languages so far, clues calibrated to CEFR levels (A1–C2)

  • Two daily puzzle sizes — a quick 7x7 and a bigger 15x15

  • Freemium; the premium tier adds full audio for every clue

  • Built-in spaced repetition, so words you've seen come back at the right intervals

I write the clues to be a bit playful or surprising rather than dictionary-dry, because that's the part that actually makes them stick — a weird or funny clue lodges in your memory in a way a definition never does.

It's live now, and this is me starting to build in public. The thing I'm genuinely unsure about is distribution — I've got the product where I want it, but I'm early on figuring out which channels work for a niche-but-broad consumer app like this. If you've launched consumer (not B2B) products, I'd love to hear what actually moved the needle for your first thousand users.

crosswort.app

3 Comments

  1. 2

    The recognition vs recall distinction is a strong framing for the product. I’m curious whether users themselves describe the problem that way, or whether that’s mainly how you arrived at the idea.

    1. 1

      It was mainly from my experience, I noticed there were certain words in english that I didn't know when I started doing crosswords extensively that through the clues and answers I had memorized - so I thought why not memorize something useful instead of Brian Eno.

      1. 1

        That makes sense — I like that the idea came from your own experience with crosswords rather than a predefined learning framework. Have you seen any users naturally describe the same “I recognize it but can’t recall it” gap yet?

About

I love NYT crosswords and suck and learning languages, I built something to help vocabulary stick with productive relaxation.