Audio-to-Score turns audio into sheet music, plus a Shazam-like feature for mobile finds the scores for music you hear

MuseScore, the cross-platform sheet music platform from music tech leaders Muse Group, has introduced two innovative audio recognition features. Audio-to-Score, currently in beta, turns audio files into ready-to-play sheet music and is now available on the web desktop version, and a Shazam-like music recognition feature is live for iOS and Android app users. Both features are free to use and reflect MuseScore’s determination to make music playing and learning more accessible and fun for creators of all levels.

"At Muse Group, we are always looking for ways to make music creation easier. With these features, we wanted to capture the spark that happens the moment people hear a song and think, 'maybe I could play this.' That spark can lead to a wonderful music session – but unfortunately often leads nowhere if there are too many steps involved," says Maxim Barchukov, Product Owner, MuseScore. "Our job is to make it easy to start playing, without spending time searching for the right score. You hear a song, you have the audio, and within a minute you have a score you can play."

Audio-to-Score: turn your audio into sheet music in minutes

With the newly developed Audio-to-Score feature, users can upload an MP3 file (up to 30 MB, or about three minutes of audio), paste a YouTube or Audio.com link and receive a sheet-music file in MSCZ format, which they can then edit and save in MuseScore Studio, the free desktop composition app integrated with the MuseScore platform.

The feature is designed to serve a range of creatives: professionals can get instant notation for their own performances, teachers can quickly prepare score excerpts from audio ahead of lessons, students can play an assigned piece and get the corresponding score, and beginners can turn a song they like into sheet music in minutes and try playing it themselves.

Audio-to-Score already recognizes solo piano – from simple pop and neoclassical piano to piano ballads – and acoustic guitar. The feature is powered by an AI model, fully developed and trained in-house by MuseScore, using deep learning to turn raw audio into structured and ready to play notation. It accurately detects rhythm and time signature, and recognizes note density and dense chords. Looking ahead, the platform plans to recognize a wider range of instruments and more complex music, and add streaming capture from other platforms, making the switch from listening to creating music even more seamless.

Shazam for scores

Another new feature, designed to be used via the iOS and Android app, helps players instantly find scores for the music they've just heard. The tool is powered by Shazam's ShazamKit technology, but instead of just identifying the song, it surfaces matching scores from the MuseScore library, which includes millions of community and licensed scores. The recognition button lives in the search bar, allowing users to seamlessly switch between manual search and music recognition.

The feature engages creators and encourages them to pick up an instrument, to master a favourite song or try something new without spending time searching for scores. The results can be saved to the library, so players can return to the scores they just found. If the tool can't find scores for a specific song, it will offer other scores available from the artist.

‍

Together with Audio-to-Score, these features help further build the MuseScore ecosystem as an unmissable destination for MuseScore’s 44 million yearly players and composers, as both playing sheet music and composing sheet music become more accessible. Earlier this year, the composition app MuseScore Studio released its latest version 4.7, adding new guitar notation along with new engraving tools and workflow upgrades.