1. Set filters
Choose genres, countries, and decades—or leave every filter open.
Discover real existing songs at random — not AI-generated music.
Combine optional filters, generate 1, 5, 10, or 20 songs, and turn no-repeat mode on or off.
Choose genres, countries, and decades—or leave every filter open.
Request 1, 5, 10, or 20 songs in each queue.
Turn on no-repeat mode to keep previously shown songs out until the matching pool is exhausted.
Open YouTube or Spotify, save favorites locally, or copy the queue.
The current catalog contains 3,731 real songs from 70 countries, released from 1944 through 2026. It combines a curated seed list with a static Wikidata CC0 snapshot updated on 2026-11-01 UTC. After filters are applied, Fisher–Yates shuffling samples the eligible catalog uniformly without popularity weighting. A total of 2,626 entries include validated YouTube video IDs.
No. It selects real, existing songs from a disclosed catalog.
Each eligible catalog entry has the same inclusion probability. Popularity is not used as a weight.
Combine genre, country, and decade, then generate 1, 5, 10, or 20 results.
Tracks open through YouTube embeds or YouTube and Spotify searches.