Google Lyria 3: Free AI Music Generator Launches in Gemini

Google launched Lyria 3, its most advanced AI music generation model, directly inside the Gemini app on February 18, 2026. The tool creates 30-second tracks with vocals, instrumentals, and auto-generated lyrics from simple text prompts, images, or videos - and it’s free for all users 18 and older.

How Lyria 3 Works

Lyria 3 lives inside the Gemini app at gemini.google.com. Users type a prompt describing the music they want, and the model generates a complete 30-second track in seconds.

The model accepts three input types:

• Text prompts - Describe the genre, mood, tempo, instruments, and vocal style
• Images - Upload a photo and Lyria 3 creates music to match its mood
• Videos - Feed a clip and get a custom soundtrack

Each generated track comes with auto-generated lyrics (unless you request instrumental-only) and a custom cover art image created by Google’s Nano Banana model.

Prompt Control

Google’s prompt guide reveals granular control over output. Users can specify vocal styles (male, female, baritone, soprano, breathy, gravelly), genre, tempo, dynamics, and instrumentation. Providing your own lyrics is as simple as prefixing them with Lyrics: in the prompt.

Example prompts from Google:

• “A comical R&B slow jam about a sock finding their match”
• “Use these photos to create a track about my dog Duncan on a hike in the woods”
• “Create a fun afrobeat track about childhood memories and home cooked plantains”

Key Capabilities

SynthID Watermarking

Every track generated by Lyria 3 is embedded with SynthID, Google DeepMind’s imperceptible watermarking technology. The watermark survives common audio edits including noise addition, MP3 compression, and speed changes.

Users can verify whether a track was AI-generated by uploading it to Gemini and asking if it was created with Google AI. Google also offers a SynthID Detector verification portal with early access for journalists and media organizations.

Availability and Pricing

Lyria 3 is free for all Gemini users 18 and older, with a global rollout that started on desktop (web) on February 18, 2026. Mobile app access is rolling out over the following days.

Google AI Plus, Pro, and Ultra subscribers get higher usage limits, though exact limits have not been published.

Supported Languages

The feature launched in eight languages: English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese, with more planned.

How Lyria 3 Compares to Suno and Udio

Lyria 3 occupies a different space than Suno and Udio. Where Suno targets musicians and producers with full-length songs up to 8 minutes and multi-stem export, and Udio focuses on polished studio-quality tracks, Google positions Lyria 3 as a casual, accessible tool for short-form content.

The 30-second cap limits Lyria 3 to jingles, social media soundtracks, and creative experiments rather than full song production. However, the free access and multimodal input - especially image and video prompts - give it a unique edge for content creators who need quick, custom audio.

Google’s AI Music Timeline

Google’s path to Lyria 3 spans three years of music AI development:

• 2023 - MusicLM, Google’s first text-to-music research model, demonstrated high-fidelity generation from text descriptions
• 2023 - Original Lyria model launched, powering YouTube Dream Track with AI voice clones of artists like Charlie Puth and T-Pain
• 2024 - MusicFX launched in Google Labs, enabling text-to-music tracks up to ~70 seconds with over 10 million tracks created
• 2025 - Lyria 2 debuted, powering YouTube’s “Speech to Song” tool
• 2026 - Lyria 3 launches in Gemini with multimodal inputs, auto-generated lyrics, and global free access

What This Means

For Content Creators

Free AI music generation inside an app billions of people already use changes the calculus for short-form content. YouTube Shorts, Instagram Reels, and TikTok creators can generate custom soundtracks without leaving Gemini or paying for a separate music subscription.

For the Music Industry

The launch reignites copyright concerns that have followed AI music since Suno and Udio faced lawsuits in 2024. Google claims it was “very mindful of copyright and partner agreements” when training Lyria 3, but has not disclosed its training data. Music Business Worldwide reports the model uses content Google “has the right to use” under its Terms of Service and partner agreements, including the UMG-YouTube deal from October 2025 that included guardrails for AI content.

For Competitors

Suno and Udio remain superior for serious music production with longer tracks, stem export, and commercial licensing. But Google offering free AI music to its massive user base could compress the casual end of the market. Royalty-free music tools like Soundraw face new pressure from a zero-cost alternative backed by Google’s distribution reach.

FAQ

---

Sources

1. Google Blog - Lyria 3 Announcement - February 18, 2026
2. Google DeepMind - Lyria Model Family - February 2026
3. Ars Technica - Gemini Can Now Generate AI Music - February 18, 2026
4. TechCrunch - Google Adds Music Generation to Gemini - February 18, 2026
5. Music Business Worldwide - Lyria 3 Launch - February 2026
6. The Verge - Google Gemini AI Music Maker - February 2026
7. 9to5Google - Gemini App Music Lyria 3 - February 18, 2026

---