impossible → routine
A whole song, generated
Generating a complete song, music and singing, as audio.
Impossible
OpenAI's Jukebox generates music with singing in the raw audio domain, at about nine hours of sampling per minute of song.
about 9 hours of compute per minute of audio
source3 years, 10 months
Routine
Suno's v3 model makes full two-minute songs in seconds and is available to all users.
a two-minute song in seconds
sourceThe Jukebox paper gives its figure in two parts: around an hour to generate one minute of top-level tokens, and around eight hours to upsample that minute to finished audio. The impossible end is dated to the paper's arXiv submission, 30 April 2020.