On Wednesday, Google revealed it is introducing a music generation capability to its Gemini app. This new feature, currently in beta, is powered by DeepMind's Lyria 3 music generation model. Users can create a song by simply describing it; the app will then produce a track complete with lyrics. For example, a prompt for a "comical R&B slow jam about a sock finding its match" would result in a 30-second song and accompanying cover art generated by Nano Banana. Google also noted that users can upload a photo or video, and the AI tool will compose a song to fit the mood of that media file. Loading the player…

According to the company, Lyria 3 represents an advancement over prior models, delivering music that is more realistic and complex. Users have additional control, allowing them to adjust elements such as style, vocals, and tempo. Beyond its integration into Gemini, Google is also providing the Lyria 3 model to YouTube creators globally via the Dream Track feature on YouTube, a tool for crafting AI-generated tracks. Previously, this option was limited to creators in the U.S.

Google clarified that the tool is not designed to directly imitate an artist. However, if a user includes an artist's name in their prompt, Gemini will produce a track inspired by that artist's general style or mood. (It remains uncertain whether this process might make it simpler for others to analyze and decode a specific artist's musical style.)

In a blog post, the company stated, “Music generation with Lyria 3 is designed for original expression, not for mimicking existing artists. If your prompt names a specific artist, Gemini will take this as broad creative inspiration and create a track that shares a similar style or mood. We also have filters in place to check outputs against existing content.”

Google also highlighted that all music created with the Lyria 3 model will include a SynthID watermark to identify it as AI-generated. Furthermore, new capabilities are being added within Gemini to detect AI-generated music using SynthID, allowing users to upload tracks and ask the app if they were AI-produced.

The music generation feature is now rolling out to all Gemini users aged 18 and older worldwide, with support for English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese.

The rise of AI-generated music has elicited mixed reactions from artists and audiences. While platforms such as YouTube and Spotify are embracing AI and entering agreements with music labels to commercialize AI-created music, AI model and tooling companies are simultaneously facing copyright lawsuits from the music industry over their training materials. In response, services like Deezer have released tools to label AI-generated music, aiming to reduce fraudulent streaming of such content.