Lyria / MusicFX logo

Lyria / MusicFX

voice

by google

Google DeepMind's AI music generation platform, from text to track with vocals and lyrics, plus a real time DJ experience.

Key features

Text to Track with Vocals and Lyrics (Lyria 3)
Image and Video to Soundtrack
MusicFX DJ: Real Time Streaming
Weighted Multi Prompt Blending
BPM and Key Control
Lyria RealTime API for Developers
Pricing

Free tier available, Google AI Pro at $19.99/mo

Best For

Creators wanting AI generated music with vocals and lyrics, not just instrumental tracks

Verdict

Free tier available via MusicFX experiments and basic Gemini access, with no credit card required to start

What it does

Text to Track with Vocals and Lyrics (Lyria 3)

Describe the music you want in natural language and Lyria 3 generates a complete track with vocals, lyrics, and instrumentation. Available in the Gemini app. A major step beyond instrumental only generation.

Learn more

Image and Video to Soundtrack

Upload an image or video and Lyria 3 generates a matching soundtrack based on the visual mood, content, and style. Useful for content creators who need music that complements their visual material.

Learn more

MusicFX DJ: Real Time Streaming

A live music generation experience where you steer the music in real time by adjusting weighted prompts, BPM, and key. Music streams continuously at 48kHz stereo. Designed with Jacob Collier.

Learn more

Weighted Multi Prompt Blending

In MusicFX DJ, blend multiple text prompts simultaneously with adjustable weights to create unique hybrid sounds. For example, blend 70% jazz piano with 30% electronic beats and adjust the mix in real time.

Learn more

BPM and Key Control

Set the tempo (BPM) and musical key for generated tracks, giving you parametric control over the musical output beyond what text prompts alone can achieve.

Learn more

Lyria RealTime API for Developers

The Lyria RealTime model provides a developer accessible API for building real time music generation into applications, games, and interactive experiences.

SynthID Watermarking

All music generated by Lyria models is embedded with SynthID, Google DeepMind's imperceptible digital watermark. This enables identification of AI generated audio without affecting the listening experience.

Cover Art Generation (Nano Banana)

MusicFX DJ automatically generates cover art for your created tracks using Nano Banana, Google's image generation model, creating a complete audio and visual package.

Multi Language Vocals

Lyria 3 can generate vocals in multiple languages, expanding the creative possibilities beyond English only music generation.

Learn more

Genre Diversity

Generate music across a wide range of genres: pop, funk, jazz, classical, drum and bass, ambient, hip hop, rock, electronic, lo fi, and many more. The model understands genre conventions and can blend styles.

YouTube Dream Track

An integration for YouTube Shorts creators that lets them generate background music and soundtracks for their short form video content using Lyria models.

Pricing

Free

Free

MusicFX and MusicFX DJ are free experiments on labs.google. Basic Gemini access includes Lyria 3 with 50 daily AI credits.

  • MusicFX text to music (experimental, free)
  • MusicFX DJ real time streaming (experimental, free)
  • Basic Gemini access with Lyria 3
  • 50 daily AI credits
  • SynthID watermarked output
  • Cover art generation

Google AI Plus

$7.99/month

Higher Gemini limits for Lyria 3 music generation with 200 credits per month. Access to text to track with vocals and lyrics.

  • Lyria 3 in Gemini (vocals and lyrics)
  • 200 credits per month
  • MusicFX and MusicFX DJ (free experiments)
  • Image/video to soundtrack
  • Multi language vocals
  • SynthID watermarked output
Best Value

Google AI Pro

$19.99/month

Generous music generation limits with 1,000 credits per month. Higher model access and longer sessions.

  • Lyria 3 in Gemini (vocals and lyrics)
  • 1,000 credits per month
  • MusicFX and MusicFX DJ (free experiments)
  • Image/video to soundtrack
  • Multi language vocals
  • Higher rate limits
  • SynthID watermarked output
Best value

Google AI Ultra

$249.99/month

Maximum music generation capacity with 25,000 credits per month. Highest model limits and priority access.

  • Lyria 3 in Gemini (vocals and lyrics)
  • 25,000 credits per month
  • MusicFX and MusicFX DJ (free experiments)
  • Image/video to soundtrack
  • Multi language vocals
  • Highest rate limits and priority access
  • SynthID watermarked output

Pros & Cons

Pros

  • Free tier available via MusicFX experiments and basic Gemini access, with no credit card required to start
  • Lyria 3 generates complete tracks with vocals and lyrics, a significant leap beyond instrumental only AI music
  • MusicFX DJ provides a unique real time music generation and steering experience that no competitor currently matches
  • Enterprise API available via Vertex AI (Lyria 2) for developers building music features into applications
  • SynthID watermarking enables responsible AI by making generated music identifiable without affecting audio quality
  • Image and video to soundtrack capability is particularly useful for content creators who need matched audio
  • Designed with input from world class musicians (Jacob Collier collaboration on MusicFX DJ)

Cons

  • 30 second clip limit in the Gemini app, which is short for many creative applications
  • MusicFX and MusicFX DJ are experimental products that could change, evolve, or be discontinued
  • Cannot mimic specific artists or generate music in the style of named performers due to policy restrictions
  • MusicFX DJ is currently only available in the United States
  • Positioned for personal creative expression rather than professional music production workflows
  • Lyria 3 is not yet available on the Vertex AI API; enterprise developers are limited to Lyria 2 (instrumental only)
  • Credit based access in Gemini means generation capacity is limited and shared with other Gemini AI features

How to get started

1

Try MusicFX for free

Visit MusicFX on Google Labs. No account or subscription required. Type a text prompt describing the music you want (genre, mood, instruments, tempo) and generate a track. This is the fastest way to experience AI music generation.

2

Experience MusicFX DJ

Open MusicFX DJ on Google Labs (US only). Enter multiple text prompts, adjust the weights between them, set BPM and key, and listen as music streams in real time. Move the sliders to steer the music as it plays. Download 60 second clips of your creations.

3

Generate tracks with vocals in Gemini

Open the Gemini app and ask it to create a song. Describe the genre, mood, lyrics theme, and vocal style. Lyria 3 generates a complete track with vocals and lyrics. You can also upload an image or video and ask Gemini to create a matching soundtrack.

4

Build with the Vertex AI API (developers)

For developers wanting to integrate music generation into applications, the Lyria 2 model is available on Vertex AI at $0.06 per 30 seconds of generated audio. The API generates 32.8 second instrumental clips. Review the documentation for integration details.

Deep dive

Detailed guides with comparisons, tips, and visuals for each feature.

Get notified about updates

We'll email you when this tool's pricing or features change.

Last updated: 2026-02-21