A shadow library group has compiled an almost complete copy of Spotify’s streaming catalog, marking one of the largest music-related data leaks ever and shedding new light on the platform’s content and listener patterns.
Known as Anna’s Archive, the group recently put together a massive pirate collection: about 86 million audio files and metadata for 256 million tracks. They say that adds up to almost everything Spotify had streamed by July 2025, with their numbers showing 99.9% catalog coverage and audio files accounting for 99.6% of all streams.
The dataset is currently being shared as a torrent approaching 300TB and includes the full metadata collection. Anna’s Archive says it plans to add the actual music files later, along with album artwork and the data required to recreate the original tracks. Most of the audio is encoded in 160kbps OGG, while less popular songs appear only in lower-quality 70kbps versions.
The activist group, which describes itself as a digital preservation project inspired by shadow libraries such as Z-Library, said it released the data as a single bulk torrent for archival reasons. It added that it may allow individual song downloads at a later date.
Spotify acknowledged a security incident to several media outlets but declined to outline its full extent. The company said it has since banned the account tied to the scraping activity and shut down the technical loophole used by Anna’s Archive.
The data offers a rare peek inside Spotify’s library, and one takeaway stands out: most of the music simply isn’t being heard. The group says at least 70% of songs on the platform have zero streams, and that the top three tracks outperform tens of millions of songs at the very bottom put together.
The archive points to the dominance of singles, the prevalence of multiple versions of the same songs, and an estimated 34 million tracks labeled as explicit.
Another big finding is just how much AI-generated music has made its way onto the platform. Anna’s Archive says a sizable chunk of the content it pulled looks to be AI-created, a sign that the technology is already reshaping Spotify’s landscape.
This growth may explain the rapid increase in album releases on Spotify, from about 2 million in 2019 to more than 10 million in 2023. The dataset also suggests that the 86 million scraped audio files, covering nearly all streams, make up only around one-third of Spotify’s total catalog, with most tracks drawing little to no listener attention.
For now, Anna’s Archive offers a controversial, unauthorized snapshot of today’s streaming economy, one dominated by a small number of hits while tens of millions of tracks go unheard.
Maybe you would like other interesting articles?