Google Gemini: Taking another big step in the world of Artificial Intelligence, Google DeepMind has launched its new music generation model Lyria 3. According to the company, this advanced AI model will now be available for Android and iOS users through Gemini app. Its purpose is to create great music tracks with simple text instructions.
Lyria 3 is designed specifically for creators who want to quickly create professional-sounding music for social media, video or digital storytelling. It is based on multimodal generative AI technology, that is, it can generate audio by understanding different inputs like text, image and video.
Complete song will be prepared from text
Introducing Lyria 3, our new music generation model in Gemini that lets you turn any idea, photo, or video into a high-fidelity track with custom lyrics.
From funny jingles to lo-fi beats, you can create custom 30-second soundtracks for any moment.
See how it works. 🧵
— Google Gemini (@GeminiApp) February 18, 2026
With the help of Lyria 3 inside the Gemini app, users can create original music tracks of up to 30 seconds. For this, you just have to tell what kind of genre is required, what should be the mood, what should be the speed of the music or which instruments should be included. The model interprets these instructions and creates a finished audio clip that includes background music, layered instruments, and balanced sound output.
With more smartness and control than before
The company claims that Lyria 3 is more advanced than its earlier version. Now it can also create lyrics automatically based on the given prompts. Also, the user gets more control over aspects like style, vocal tone and tempo. Better sound quality and ability to maintain harmony even in long compositions has been added to it.
Soundtrack will also be made from photos and videos
Lyria 3 isn’t just limited to text. If you upload a sunset photo or travel video, it can create music according to its atmosphere. Not only this, cover art generated by AI can also be added to each track, making it easier to share the content. Additionally, Google is also integrating this technology with YouTube Dream Track where creators will be able to experiment with different music styles for shorts and other formats.
watermarking for transparency
To ensure identification of AI-generated content, the company has included SynthID watermarking technology in Lyria 3. This adds an invisible identity to the generated audio, making it clear that it has been created by AI.
According to Google, this move is part of its broader strategy towards a multimodal AI platform, where text, images, audio and video will work in a single ecosystem. Lyria 3 is currently being made available in multiple languages for users above 18 years of age, while its expansion to the mobile app will be done in a phased manner.
Also read:
New feature of Google! Now the files in Downloads folder will also be safe, know how it will work
