Lyria 3 | Music Generation comes to Gemini

The launch of Lyria 3 marks a major shift in how you can create and interact with music directly within the Gemini ecosystem. Powered by Google DeepMind’s latest technology, it transforms text, images, and even videos into high-fidelity audio tracks.

Whether you’re looking for a quick social media loop or a structured song, here is the breakdown of what Lyria 3 brings to your workflow:

The Models: Lyria 3 vs. Lyria 3 Pro

Google has introduced two distinct tiers to balance speed with creative depth:

• Lyria 3 (Clip): Best for rapid iteration and social sharing. It generates high-quality 30-second tracks and is available to all users on the Free tier.

• Lyria 3 Pro: Designed for advanced utility. It can generate full-length songs up to 3 minutes, complete with complex musical structures like verses, choruses, and bridges. This is accessible to AI Plus, Pro, and Ultra subscribers.

Key Features & Capabilities

1. Multimodal Creation

You aren’t limited to text descriptions. You can influence the “vibe” of your music using different inputs:

• Text-to-Music: Describe genres, moods, instruments, or even a specific story.

• Image/Video-to-Music: Upload a photo or video, and Gemini will analyze the visual mood to compose a matching soundtrack.

• Automated Lyrics: Lyria 3 writes the lyrics for you based on your prompt, performing them with realistic vocal nuance.

2. Structural Control (Pro Only)

For those using the Pro model, you can use timestamped prompting to dictate exactly how a song progresses:

3. Real-Time “Steering”

Through the Gemini API and AI Studio, developers can use Lyria RealTime. This allows for a “DJ-like” experience where you can adjust parameters like BPM, density, and brightness while the music is playing, causing the track to morph smoothly in real-time.

How to Use It

1. Open Gemini: Access it via the web (gemini.google.com) or the mobile app.

2. Locate the Tool: Look for the Music icon in the toolbar or simply start your prompt with “Create a track…”

    

3. Refine Your Prompt: Be specific. Mention the BPM (e.g., “120 BPM”), the instrumentation (e.g., “cello and lo-fi beats”), and the emotional tone (“nostalgic but hopeful”).

4. Download: Once generated, you can download your track as an MP3 (or WAV for Pro users) to use in your projects.

    

Safety & Transparency

Every track generated by Lyria 3 is embedded with SynthID. This is an imperceptible digital watermark that allows you to verify if audio was AI-generated, ensuring transparency in creative work.

Are you looking to create a specific type of track, perhaps something for your training sessions or a background score for a project?