The app’s latest Lyria 3 model creates high-quality, 30-second tracks featuring vocals and lyrics from straightforward text prompts or uploaded images.
What you need to know
– Gemini now produces 30-second songs with lyrics derived from text prompts or images.
– This feature is powered by Google’s Lyria 3 model, which manages lyrics, style adjustments, and delivers more realistic audio.
– Enter a prompt or upload an image/video, and Gemini transforms it into a song ready for sharing, complete with custom cover art.
The Gemini app has expanded past just text and images. As of today, it aims to compose your next song.
Gemini has introduced music generation in beta, utilizing Lyria 3, Google DeepMind’s latest generative music model. This allows you to quickly create a 30-second track with lyrics by simply providing a prompt or even a photograph.
Google states Lyria 3 enhances previous versions in three significant ways: it can automatically generate lyrics, allows for more precise control over style, vocals, and tempo, and creates tracks that sound more authentic and layered. In simple terms, you don’t need to be a songwriter. Just convey a vibe, like “a nostalgic afrobeat homage to my mom’s cooking” or “a quirky R&B tune about a lonely sock,” and you’ll receive a refined mini-track in return.
Photo or video to track
You may also upload a photo or video, and Gemini will convert that moment into a soundtrack with complementary lyrics. For instance, a hiking picture of your dog could evolve into a woodland anthem complete with lyrics. The system even generates custom cover art using Nano Banana, ensuring your track is ready to be shared directly from the app.
Since its debut, Gemini has incorporated additional creative tools, transitioning from images to video. Music is the next phase. Nonetheless, Google isn’t attempting to turn this into a Spotify alternative or a hit-making apparatus. The goal is to create an enjoyable method for integrating music into your everyday life without requiring complex software or needing to hire a producer.
There’s also something available for creators. Lyria 3 is on its way to Dream Track on YouTube Shorts, beginning in the U.S. and expanding globally. This means you can obtain improved, customizable backing tracks and lyrics for short videos, which is crucial since audio plays a significant role in engagement.
Naturally, AI music raises copyright issues. Google asserts Lyria 3 is developed with careful consideration of copyright and partner agreements, aimed at original expression rather than replication. If you refer to a specific artist, Gemini utilizes it as inspiration rather than a direct reference. There are filters in place to identify existing content, and users have the option to report potential infringements.
Each generated track incorporates SynthID, Google’s invisible watermark for recognizing AI-generated content. Gemini’s verification tools now extend to audio as well, allowing you to upload a file and inquire whether it was produced with Google AI. The system checks for SynthID and conducts its own analysis before providing a response.
Lyria 3 is now accessible in the Gemini app for users aged 18 and older, in various languages such as English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese. It is launching on desktop first, with mobile support coming soon. Google AI Plus, Pro, and Ultra subscribers enjoy higher usage limits.
Android Central’s Take
Personally, I find this to be one of the more practical AI enhancements we’ve encountered. Not because it will substitute musicians — it won’t — but due to its capability to reduce the barriers to creative expression in a way that feels fun rather than daunting. If you’ve ever desired a custom birthday song, a theme for your group chat, or a quirky track to share on Shorts, you can now obtain one in seconds. For users, this translates to creativity on demand, which is quite an enticing direction for AI to pursue.