Rhymeprint
phonetics, not spelling
Rhymeprint maps a verse the way it actually sounds, syllable by syllable, including the chains that run straight through word boundaries. Paste the lyrics or load the song. It all runs in your browser, so nothing you write or play leaves your device.
Try it. Type or paste anything
Analyzed in your browser just now. Switch the style above. The seal is this verse's Rhymeprint: one segment per syllable, coiling outward as the verse runs.
the notation
Colour means sound. Everything else stays out of the way.
"Climate" and "grind it" carry the same colour because they share vowels, not letters. The notches show you the syllables inside the rhyme. That is the part other tools flatten.
The verse's dominant schemes sit at full strength; incidental rhymes fade back. Pick Blocks to keep one mark for everything, or Layered to drop the strays to coloured text.
If a word isn't part of a rhyme, it gets no highlight at all. A glance tells you how much of the bar is actually carrying the scheme.
Alliteration and sibilance. The consonant texture sitting under the vowels, marked quietly enough that it never fights the colour.
how it works
Paste a verse, or load the song and hit Transcribe. A speech model runs in your browser and drafts the lyrics, already timed to the track. Works on unreleased demos, because nothing is uploaded.
Every word is pronounced from a phonetic dictionary, in the accent you choose, then matched by sound. You see the multis, the internal chains, and the exact bar where the pattern turns.
Record the rhymes lighting up in time with the vocal, audio in the file, in whichever shape you post to. You can also export the still and the Rhymeprint card. It renders on your machine.
what it's for
Paste the one you cannot stop rewinding and see what it is doing underneath. Where the multis run, and the exact bar where the scheme flips.
Run your drafts. Thin colour means a thin scheme: you'll see where a chain dies a bar early and where a multi almost lands. Fix it before you record it.
No lyrics handy? Load the track and hit Transcribe. A speech model runs in your browser and drafts the words, timed to the song. Even your unreleased demo stays yours.
Load the song and Rhymeprint renders the rhymes lighting up in time with the vocal. Vertical for TikTok and Shorts, square for the feed, widescreen for YouTube.
Some rhymes only land in the delivery. Southern pulls "ride" toward "rod" and drops the g in "rollin'"; UK lets "laugh" reach "scarf". Pick the accent and the engine pronounces the verse that way.
Show a class or a comment section exactly why a bar hits. The colours do the explaining, and the export drops straight into a slide or a post.
why it's different
open beta
It already handles multis, mosaics, homographs and three accents. Try to break it, then tell me which bar it got wrong. That is what the next release gets built from.
pricing
Pro is for the clips you put out. It takes the banner off and renders at full quality.
The whole engine, no account.
For anyone posting the clips.
Cancel whenever. Nothing to uninstall.
questions
Nowhere. The engine downloads into your browser and the analysis runs on your device. There is no server for lyrics to reach. That is the architecture, not a policy.
Sounds, not spellings. Every word gets pronounced first, using dictionary phonetics plus slang handling and whichever accent you picked. Then the engine looks for recurring stressed-vowel patterns, including multisyllable chunks that run across word boundaries.
Only the drawing. Blocks uses one mark for every rhyme and fades the lesser families, so the page reads as a single visual language. Layered keeps solid highlights for the main schemes and drops stray rhymes to coloured text. The analysis underneath is identical. Switch it on the demo above and watch it flip.
Load the song and sync it. Transcribe does that part for you, or you can tap-sync the lines by hand. Then hit Make video and Rhymeprint renders the file on your machine, audio included, usually faster than the song is long. You do not have to sit and watch it happen.
It is a Whisper speech model running in your browser. Treat it as a strong draft rather than a stenographer. Clean vocals and acapellas come out close, and a dense mix needs a quick pass of fixes. It is there so you can analyze a song without typing it out, including one nobody has heard yet.
Sometimes, and that's why it's in open beta. English spelling hides its own pronunciation, so an unusual name or a heavy slang contraction can land in the wrong family. Send me the verse and the bar and I will fix it.
Yes. Exports of your analyses are yours, on Free or Pro. The lyrics you paste are your responsibility. Rhymeprint never stores them and never republishes them.
No account. Nothing uploaded. Takes about five seconds.