'Anna's Archive,' a site best known for backing up books and academic papers, has made one of its most controversial claims yet: that it has scraped almost the entirety of Spotify’s catalog.
According to a detailed blog post published by the group, the archive now contains metadata for roughly 256 million tracks and audio files for about 86 million songs. This represents around 99.6% of all listens on the platform. The full collection weighs just under 300TB and is being distributed via bulk torrents, organized by popularity.
If accurate, this would make it the largest publicly available music metadata database in existence.
Anna’s Archive describes the project as a preservation effort, arguing that while popular music is usually well backed up, vast amounts of lesser-known material are vulnerable to disappearing as streaming licenses expire, catalogs change, or platforms shut down.
In its own words, "Spotify doesn’t have all the music in the world, but it’s a great start."
The archive explains that it prioritized tracks using Spotify's own popularity metric.
For songs with nonzero popularity, the group claims it captured nearly all available audio, stored in Spotify's original streaming quality of around 160 kbps OGG Vorbis. Tracks with zero popularity were archived more selectively, often re-encoded into smaller OGG Opus files at roughly 75 kbps to reduce storage requirements.
Anything released after July 2025 may be missing or incomplete.
While metadata is already fully available, the audio files are being released gradually, starting with the most popular tracks.
Anna’s Archive estimates that it has backed up about 99.9% of Spotify’s 256 million-track library in metadata form, with roughly 186 million unique ISRCs represented. The group says the compressed metadata itself is relatively small (under 200GB) while additional datasets, such as audio analysis files, add several terabytes more.
By comparison, widely used open music projects like MusicBrainz contain only a fraction of that scale.
The group justifies the project as part of a broader mission to preserve human knowledge and culture across all media.
"Our mission (preserving humanity’s knowledge and culture) doesn’t distinguish among media types," the blog post states. "Sometimes an opportunity comes along outside of text. This is such a case."
It also argues that existing music archiving efforts suffer from an over-focus on popular artists, an emphasis on large lossless files that make comprehensive archives impractical, and the absence of a single authoritative torrent collection intended to represent all recorded music.
Spotify, however, sees the situation very differently.
In a statement provided to Billboard, a company spokesperson said, "An investigation into unauthorized access identified that a third party scraped public metadata and used illicit tactics to circumvent DRM to access some of the platform's audio files."
Spotify added, "We are actively investigating the incident," and later confirmed that it had identified and disabled accounts involved in what it described as unlawful scraping, while implementing additional safeguards against similar attacks.
Legally, the situation is straightforward even if the motivations are disputed.
Spotify licenses the vast majority of its music under strict agreements with record labels and rights holders. Mass scraping of audio files and redistributing them via torrents violates both Spotify's terms of service and copyright law in many jurisdictions. While Anna's Archive frames the project as preservation rather than piracy, copyright law generally does not provide exceptions based on intent alone.
The scale of the archive raises uncomfortable questions for the music industry and researchers alike.
As streaming services dominate music distribution, they also act as gatekeepers to cultural memory, with catalogs that shift constantly due to licensing, territorial restrictions, or artist decisions. A comprehensive snapshot of the world's largest streaming platform, especially one tied to popularity data, would be invaluable for research, but it comes wrapped in serious legal and ethical problems.
Not to mention, a 300TB dataset is enormous. Even though it’s being distributed via torrents across global peer-to-peer networks, very few people can realistically store that much data locally.
As a result, only a small number of highly committed individuals are likely to keep complete copies and continue seeding the archive, effectively carrying it forward for audiophiles and researchers who lack the storage to hold it themselves.
