github NeptuneHub/AudioMuse-AI v2.0.0
AudioMuse AI v2.0.0: Lyrics new model

latest releases: v2.0.2, v2.0.1
12 hours ago

Release Date: May 23, 2026

AudioMuse AI v2.0.0 upgrades the lyrics engine to a new multilingual model. What this means for you:

  • Better language support: lyrics are understood in their original language instead of being translated to English first.
  • Big improvement on CJK (Korean / Japanese / Chinese).
  • More lyrics analyzed per song: Now the token analyzed per song pass from 128 to 512 for more accurate results.
  • Speed: Avoiding the translation step, the speed of Lyrics step is improved.

This release also improve the Instant Playlist functionality that create playlist based on chat with AI. It is now able to better compose together the different tools instead of just merging them, for better result.

Table of functinality result of all the functionality was standardized and improved by including additioanl information like genre and moods.

Breaking change - reset the lyrics tables
Lyrics analyzed with the old model are not compatible with the new one. After upgrading, drop the lyrics tables and run a new analysis to re analyze the lyrics only on all your library.
Replace the connection parameters (ip, port and if you change db name and password) and run:

PGPASSWORD=audiomusepassword psql -h 192.168.3.213 -p 5432 -U audiomuse -d audiomusedb \
  -c "DROP TABLE IF EXISTS lyrics_embedding; DROP TABLE IF EXISTS lyrics_index_data; DROP TABLE IF EXISTS lyrics_axes_index_data;"

Or on Docker Compose (replace postgres with your DB service name):

docker compose exec -e PGPASSWORD=audiomusepassword postgres \
  psql -U audiomuse -d audiomusedb \
  -c "DROP TABLE IF EXISTS lyrics_embedding; DROP TABLE IF EXISTS lyrics_index_data; DROP TABLE IF EXISTS lyrics_axes_index_data;"

What's Changed

Full Changelog: v1.1.5...v2.0.0

Don't miss a new AudioMuse-AI release

NewReleases is sending notifications on new releases.