Udio is a generative artificial intelligence model that produces music based on simple text prompts, capable of generating both vocals and instrumentation. Developed by Uncharted Labs, a startup founded in December 2023, the service uses Generative AI techniques to create original songs from user descriptions. Its free beta version was released publicly on April 10, 2024, allowing users to generate up to 600 songs per month at no cost. The platform offers subscription tiers for advanced features such as audio inpainting, which lets users modify specific parts of a generated track.
The service emerged from a team of former researchers at Google DeepMind, positioning it within the broader wave of Artificial intelligence creative tools that also includes text-to-music competitors like Suno and Stable Audio. Udio's launch generated significant attention for its realistic vocal synthesis, while also raising questions about the ethical and legal implications of training AI models on copyrighted music.
History
Udio was created in December 2023 by a team of four former researchers for Google DeepMind, including CEO David Ding, Conor Durkan, Charlie Nash, and Yaroslav Ganin, along with Andrew Sanchez, operating under the name Uncharted Labs. The venture capital firm Andreessen Horowitz provided financial backing, as did the music distributor UnitedMasters, musicians will.i.am, Tay Keith, and Common, investor Kevin Wall, Instagram cofounder Mike Krieger, and DeepMind researcher Oriol Vinyals. The company raised $8.5 million in initial funding and was valued at $10 million in seed funding.
After several months in a closed beta phase, Udio was publicly released on April 10, 2024. Sanchez described the goal as enabling musicians to create great music and to make money from that music in the future. The release followed other text-to-music generators such as Suno and Stable Audio, marking a competitive period in AI music generation.
Udio gained viral attention when it was used to create "BBL Drizzy" by Willonius Hatcher, a parody song that circulated widely during the Drake-Kendrick Lamar feud, amassing over 23 million views on Twitter and 3.3 million streams on SoundCloud in its first week. In August 2024, "Verknallt in einen Talahon" (In Love with a Talahon), a song generated with Udio by Austrian producer Butterbro, became the first AI-generated song to reach the German Top 50.
Capabilities
Udio bases the songs it creates on text prompts, which can specify genre (including barbershop quartet, country, classical, hip hop, German pop, and hard rock), lyrics, story direction, and other artists to base the sound on. The underlying generation process, as of April 2024, had not been disclosed. The program generates two songs based on the prompts, and users can remix their songs with further text prompts. Songs are initially generated as roughly 30-second pieces and can be extended in additional 30-second increments. Paying subscribers can access advanced functionality such as audio inpainting, which allows for targeted edits within a track.
The service leverages Deep learning models, likely involving Neural network architectures similar to those used in other generative AI systems. The ability to produce realistic vocals suggests the use of sophisticated Transformer (architecture)-based models, though the company has not publicly detailed its technical approach. The user interface emphasizes simplicity, making it accessible to people with no or minimal musical ability.
Reception
Critics offered mixed assessments of Udio's output. Mark Hachman of PC World compared Udio to AI art generators and praised its ability to turn "a few rather poor lyrics" into a "rather catchy" song, calling the vocals "incredibly realistic and even emotional." Sabrina Ortiz of ZDNET described the generated songs as "impressive" and sounding as though they were professionally produced, noting they were "fuller and richer" than those of other text-to-music generators. Tom's Guide's Ryan Morrison wrote that Udio had "an uncanny ability to capture emotion in synthetic vocals" and was the only AI music generator to capture the passion, pain, and spirit of a vocal performance.
Brian Hiatt of Rolling Stone wrote that Udio was "more customizable but also perhaps less intuitive to use" than Suno, adding that some early users suggested Udio's output may sound crisper. However, Benj Edwards of Ars Technica found Udio's generation capability imperfect and "less impressive" than Suno's, noting its songs were substantially shorter and calling them "half-baked and almost nightmarish." In response to the beta announcement on Twitter, Telefon Tel Aviv member Joshua Eustis called Udio "an app to replace musicians" and questioned the data it used. Udio has also been criticized online as "soulless" and for its potential to create audio deepfakes. Lucas Ropek of Gizmodo stated that Udio was "full of acoustical nonsense" and that its songs were "extraordinarily bad."
Copyright Concerns
Critics questioned what data was used to train Udio and whether it consisted of copyrighted music. Rolling Stone wrote that there was "substantial reason to believe" that both Udio and Suno were trained with copyrighted music, while Benj Edwards of Ars Technica wrote that its training data was "likely filled with copyrighted material." Udio does not directly recreate copyrighted songs if prompted. Ding has stated that Udio has "extensive automated copyright filters" and that the company is continually refining its safeguards. Stability AI took a different approach with Stable Audio 2.0, using an explicitly licensed dataset called AudioSparx.
In June 2024, a lawsuit led by the Recording Industry Association of America was filed against Udio and Suno, alleging widespread infringement of copyrighted sound recordings. The lawsuit sought to bar the companies from training on copyrighted music and demanded damages of up to $150,000 per work for infringements that had already taken place.
In October 2025, a copyright lawsuit from Universal Music Group was settled with licensing agreements. Udio agreed to launch a platform trained only with "authorized and licensed music." The company then announced that existing users would have 48 hours to download their creations before transitioning to a new streaming-based business model.