Suno is an AI music generation system, developed by the Cambridge, Massachusetts-based startup Suno Inc., that produces complete songs, including sung vocals, lyrics, and full instrumental arrangements, from a short text description of a genre, mood, or theme. Launched publicly in December 2023, it is one of the best-known tools in the AI music generation category and, alongside rival Udio, was central to a wave of AI-generated music that began circulating widely online through 2024 and 2025.
History
Suno was founded in 2022 by a team of former audio and speech engineers, some with backgrounds at companies including Kensho and TrueAccord, and released its first consumer-facing product in late 2023. A partnership with Microsoft integrated an early version of the technology into Bing's Copilot chat experience before Suno's standalone app grew into its primary product. Successive versions through 2024 substantially improved vocal realism, lyrical coherence, and mixing quality; by version 4, released in late 2024, AI-generated Suno tracks were regularly mistaken for human-produced songs in informal listening tests, and the company reported tens of millions of songs generated by its user base. Suno raised significant venture funding during this period, reflecting broader investor interest in the AI boom (2022-present) in generative media.
Technology
Suno has not published full technical details of its architecture, but the system combines language modeling for lyric and structure generation with generative audio modeling to render the actual waveform, likely drawing on techniques related to Diffusion models or Autoregressive model audio generation used elsewhere in generative audio research. A user's Prompt can specify genre, instrumentation, tempo, and lyrical themes, or supply full custom lyrics for the model to set to music, and the output is a complete stereo audio track rather than a symbolic score or set of stems, though later versions added limited stem separation and editing tools.
Reception and legal challenges
Suno was widely adopted by hobbyists, independent musicians, and content creators for its low barrier to producing usable music quickly, and some tracks generated with it charted on informal internet music rankings or were used in viral social-media content. It also became a focal point of the recording industry's response to generative AI: in June 2024, the major record labels, through the Recording Industry Association of America (RIAA), filed a copyright infringement lawsuit against both Suno and Udio, alleging the companies trained their models on copyrighted sound recordings without a license, a case widely covered as one of the most consequential of the broader wave of AI copyright lawsuits because it directly implicated the major labels' catalogs as Training data. Suno acknowledged training on copyrighted material but argued the practice fell under fair use, a defense central to unresolved questions in AI and copyright law across generative media. The litigation, still unresolved as of 2025, has been widely watched as a bellwether for how courts will treat AI music generation and for what licensing arrangements, if any, AI music companies may eventually strike with rights holders.