Back to browse
Enunciate – Paste a speech script, find the words you'll mispronounce

Enunciate – Paste a speech script, find the words you'll mispronounce

by eventual_consis·Feb 13, 2026·1 point·0 comments

AI Analysis

●●SolidNiche GemSolve My Problem
The Take

Clever and focused: paste your script and Enunciate highlights risky words from a curated list (~350) using pattern heuristics (silent letters, odd spellings) and lets you click any word to hear a pronunciation and see IPA or simplified respelling. The small UX wins — presentation mode and US/UK toggles — make rehearsals smoother, but the approach is rule-based and limited by its wordlist; it doesn't adapt to your personal mispronunciations or contextual pronunciation edge cases.

Category
Target Audience

Presenters, public speakers, podcasters, salespeople, and anyone rehearsing spoken scripts or demos

Post Description

Hey HN,

I built Enunciate because I kept stumbling over words like "epitome" (ih-PIT-uh-mee) and "unequivocal" (un-ih-KWIV-uh-kul) during presentations. I’d rehearse my delivery, but not the specific words most likely to trip me up, and looking up pronunciations mid-rehearsal broke my flow. I couldn’t find a simple script focused tool for this, so I built one.

Enunciate lets you paste a speech or presentation script and automatically highlights words you're likely to mispronounce. Click any flagged word to hear it spoken and see its phonetic spelling. Presentation mode is designed for recordings and live demos.

How it works: - Paste your script (handles most rich text from Google Docs/Word) - Uses a curated list of commonly mispronounced words (~350) - Scores words based on patterns that tend to cause mistakes (silent letters, uncommon spellings, rare words, length) - Click a word to hear pronunciation + see phonetics (IPA or simplified respelling) - Presentation mode shows inline pronunciations

Technical details: - No custom backend — scripts are analyzed locally in the browser - Individual words are sent to dictionaryapi.dev to fetch pronunciation data - Built with React + Vite + Tailwind

I mainly built this for rehearsing talks, interviews, lectures and demos without breaking flow to Google pronunciations. Presentation mode is useful for creators as well (recording for youtube/podcasts).

Would love feedback, especially from anyone who does public speaking regularly.

Similar Projects