Lyria 3.5 AI music generator for richer, full-length tracks

Lyria 3.5 is Google DeepMind’s latest AI music generation model, built to create richer and more expressive tracks from simple creative direction. Describe the genre, mood, instruments, vocal style, tempo, or duration you want, and the model turns those details into a cohesive piece of music. Compared with earlier versions, Lyria 3.5 offers stronger musicality, more natural vocals, better lyric generation, and greater control over how a track develops. It can create everything from quick musical clips to full three-minute songs, helping creators produce audio that better fits the length and mood of their projects.

Lyria 3.5 gives you more control over how your music sounds, feels, and develops. Create richer compositions, stronger lyrics, and more expressive vocals, while adjusting the tempo and duration to fit your project. 

• Richer musicality 

  Create natural, complex melodies and cohesive arrangements that flow from start to finish.

• Stronger lyric generation 

  Generate higher-quality lyrics that better follow your prompt and intended song structure.

• More expressive vocals 

  Produce realistic, expressive vocals with nuanced delivery and clearer pronunciation.

• Greater creative control 

  Set the tempo and generate tracks at your chosen duration, up to 3 minutes.

   

Lyria 3.5 also works from images, translating their mood and atmosphere into high-fidelity music. Its range across genres and languages allows the same visual starting point to develop in different musical directions for different projects and audiences.

Lyria 3.5 is available in Picsart AI Playground, giving you direct access to Google DeepMind’s latest music generation model in the same creative workspace as Picsart’s image and video tools. Describe the genre, mood, instruments, vocals, tempo, and track length you want, then generate high-fidelity music tailored to your creative direction.