Udio launched in 2024, founded by former DeepMind and Google researchers, and rapidly became the other major name in full-song AI generation alongside Suno, known for its own take on vocal realism and genre range from a text prompt.
Major labels sued Udio the same year over training data, and both Universal Music Group and Warner Music reached settlements in late 2025, but Udio's deal came with real product changes attached: rather than continuing as an open text-to-song generator, Udio has been shifting toward a more restricted platform built around licensed music, with creations staying inside the platform rather than freely exporting the way earlier generations could. UMG has described a co-launched licensed platform arriving in 2026.
Pricing and free-tier details are in flux alongside that transition; plans have started around $10 a month historically. For anyone evaluating Udio today, it's worth confirming current terms directly, since the product settling into a licensed, more contained platform is a meaningfully different thing than the open generator it launched as.
