Gemini App Gets Musical with Lyria 3: Create Custom Tracks from Text and Images

Create custom 30-second music tracks in Gemini using Google's Lyria 3 model. Generate songs from text or images with integrated SynthID watermarking.

by HowAIWorks Team
Google GeminiLyria 3Generative AIMusic AIDeepMindArtificial Intelligence

Introduction

Music has always been a powerful form of human expression. Now, with the latest update to the Gemini app, you can turn your ideas into custom soundtracks in seconds. Google has officially rolled out Lyria 3, Google DeepMind’s most advanced generative music model, directly within the Gemini experience.

Whether you're looking to create a catchy tune for a social media post, a fun song for a friend, or just want to experiment with AI-generated music, Lyria 3 makes it incredibly accessible. Available now on desktop and rolling out to mobile devices, this new feature lets you generate high-quality, 30-second tracks complete with vocals and lyrics, all from a simple text prompt or an uploaded image.

What is Lyria 3?

Lyria 3 represents a significant leap forward in AI music generation. Building upon previous models, it offers enhanced capabilities that allow for more complex and realistic musical compositions.

Key improvements in Lyria 3 include:

  • Integrated Lyrics Generation: No need to write your own verses; the model can generate lyrics that fit your prompt's theme and mood.
  • Creative Control: Users have more influence over style, vocals, and tempo, allowing for a more personalized output.
  • Musical Complexity: The model can handle more intricate musical structures, resulting in tracks that sound more professional and less like typical AI-generated loops.

How to Create Music in Gemini

Creating music with Lyria 3 in Gemini is designed to be intuitive and fun. Here are the two main ways you can interact with the new feature:

Text to Track

Simply describe what you want to hear. You can specify a genre, a mood, a specific instrument, or even a scenario.

Example Prompt: "I’m feeling nostalgic. Create a track for my mother about the great times we had as kids and the memories of her home-cooked plantains. Make it a fun afrobeat track with a true African vibe."

Visuals to Track

If you can't find the words, let your photos or videos do the talking. Upload a visual file, and Gemini will analyze the content to compose a track that matches the vibe.

Example Prompt: "Use these photos to create a track about my dog Duncan on a hike in the woods."

Custom Cover Art

Every track you generate comes with its own custom cover art, created by a model called Nano Banana. This adds a nice visual touch to your audio creation, making it ready to share with friends or on social media immediately.

Safety and Responsibility

As generative AI becomes more powerful, responsible development is crucial. Google has implemented several safeguards with Lyria 3:

  • SynthID Watermarking: All audio generated by Lyria 3 is embedded with SynthID, an imperceptible watermark that allows tools to identify the content as AI-generated.
  • Audio Verification: New tools in the Gemini app allow users to upload audio files to check if they were created using Google's AI.
  • Artist Protections: The model is designed for original expression. If a prompt specifically requests to mimic a real artist, Gemini will use it as "broad creative inspiration" rather than producing a direct copy.

For Creators: YouTube Shorts Integration

For content creators, Lyria 3 is also making its way to YouTube via Dream Track. This integration allows creators to generate unique soundtracks for their Shorts. Whether you need a specific background vibe or a lyrical verse to punch up a video, this tool offers a new level of customization for short-form video content.

Conclusion

The integration of Lyria 3 into the Gemini app marks another exciting step in democratizing creative tools. By allowing anyone to generate custom music from text or images, Google is opening up new possibilities for personal expression and content creation.

Ready to start making music? You can try out Lyria 3 today at gemini.google.com/music. As always, remember to use these tools responsibly and have fun exploring the new sonic landscapes you can create!

Frequently Asked Questions

Lyria 3 is Google DeepMind's latest generative music model, capable of creating high-quality music with vocals and lyrics from text or image prompts.
You can use Lyria 3 in the Gemini app to generate 30-second music tracks by describing your idea in text or uploading an image or video for inspiration.
Yes, all tracks generated in the Gemini app are embedded with SynthID, an imperceptible watermark for identifying Google AI-generated content.

Continue Your AI Journey

Explore our lessons and glossary to deepen your understanding.