The Gemini chatbot now has a function for creating short music tracks: just describe the idea in words or add a photo or video, and AI will generate music and lyrics and provide a file for download or sharing.
The Gemini chatbot now has a function for creating short music tracks: just describe the idea in words or add a photo or video, and AI will generate music and lyrics and provide a file for download or sharing.
As Google explains in the instructions, the new feature works on the Lyria model and is currently being gradually rolled out to the Gemini mobile app, so it may not be available to everyone immediately. The result is sent as a video with a personalized cover, and the length of the songs is limited to 30 seconds at this stage.
There are several requirements for using music generation: the user must be 18+, logged into a personal Google account or a supported company or educational institution account, and enable saving of activity history. After creating a track, it can be downloaded as an MP3 or as an MP4 with cover art, and also share a link. Google separately notes that each composition will have a built-in digital watermark to identify the content as AI-generated.
In query suggestions, the company recommends starting with the verbs «Write,» «Compose,» or «Create,» specifying the genre and instruments, and adding details about the mood, tempo, and plot of the song. At the same time, there will be limits on the number of generations: a notification will appear after exceeding, and a Google AI subscription provides higher limits.
The editors of dev.ua also tested the new product in practice and asked AI to generate a track-"news» about exoskeletons that Nova Poshta is testing. The result was a bit corny, but musical:
