Google DeepMind has launched Lyria 3, its latest artificial intelligence music-generation model. The company has integrated the tool into the Gemini app on Android and iOS. Now, users can create high-quality music tracks simply by typing text prompts.
Moreover, the model uses multimodal generative AI, allowing creators to produce soundtracks for social media, videos, and digital storytelling with minimal effort. As a result, music production becomes faster and more accessible.
Create Music Using Simple Prompts
According to Google DeepMind, Lyria 3 powers new music-creation features inside Gemini. Users can generate original tracks up to 30 seconds long by describing genre, mood, tempo, or instruments.
The AI then converts the written prompt into a fully produced audio clip with layered instruments and polished sound. Additionally, the system automatically balances different sound elements to deliver a ready-to-use track.
Furthermore, Lyria 3 can also generate lyrics based on the user’s description. It offers better control over vocals, style, and tempo. Therefore, creators can fine-tune the output according to their needs.
Supports Photos and Videos
Besides text prompts, Lyria 3 also works with visual input. For example, users can upload a sunset photo or a short travel video. The AI analyzes the visual mood and then composes matching background music.
In addition, each generated track can include AI-created cover art. This makes it easier for creators to publish or share their work across platforms.
Integration With YouTube Dream Track
Google is also expanding Lyria 3 to YouTube Dream Track. Consequently, creators can experiment with AI-generated music for Shorts and other formats. They can quickly test multiple styles, themes, and moods without traditional music tools.
However, AI-generated content raises authenticity concerns. To address this, Google has added SynthID watermarking. The system embeds invisible metadata into the audio so platforms can identify it as AI-generated content.
Availability and Rollout
Lyria 3 forms part of Google’s broader multimodal AI strategy, where text, image, audio, and video tools operate within one platform.
Currently, the feature is available on the Gemini desktop app for users aged 18 and above. It supports English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese.
Meanwhile, the Android and iOS rollout will happen over the next few days. Additionally, Google AI Plus, Pro, and Ultra subscribers will receive higher usage limits.