Google has released Lyria 3.5, its new music generation model, with a focus on making generated tracks easier to refine after the first version is created. The update is available through Google Flow Music and centers on more natural output, tighter control, and section-level editing.
What Lyria 3.5 changes for music generation
According to Google, Lyria 3.5 produces more natural-sounding melodies, better lyrics, and more realistic vocals with clearer pronunciation. Those changes matter because generated music often depends not only on the idea behind a prompt, but also on whether the result sounds coherent enough to keep working with.
The model also gives users more precise control over tempo and track length. Tracks can run from 30 seconds to 3 minutes, giving creators room to make short pieces or longer song drafts inside that range.
That combination points to a practical shift in how the tool is meant to be used. Instead of treating the first generated track as a finished take or a throwaway attempt, Lyria 3.5 gives users more ways to shape the output toward a specific length, pace, and arrangement.
Section editing reduces the need to start over
The most important workflow change is a new feature called "Selective Section Painting". It lets users edit specific parts of a track or turn short melodies into full songs without starting over.
That is a significant usability change for anyone working iteratively. If one part of a generated track works and another does not, the user no longer has to discard the whole piece just to improve a section.
In plain terms, the feature supports a more targeted editing process. A user can focus on a particular passage, expand a shorter melodic idea, or continue developing a track while preserving the parts that already fit the goal.
The source also says users can fine-tune the tempo and duration of vocals, drums, bass, and other elements. That matters because a track is not only a single audio block; it is made from parts that need to fit together in timing and structure.
More control over vocals, drums, bass, and timing
Lyria 3.5’s controls extend beyond overall track duration. Users can adjust the tempo and duration of vocals, drums, bass, and other elements, which gives them a more detailed way to shape how a song develops.
For generated music, that kind of control can be the difference between a rough idea and something more usable. A vocal line may need to sit differently against drums, or a bass part may need to match the intended pace more closely.
The update does not simply promise better audio quality. It also addresses the editing problem that comes after generation: how to keep the useful parts of a track while changing the parts that need more work.
- Lyria 3.5 is available through Google Flow Music.
- Tracks can run from 30 seconds to 3 minutes.
- Users can control tempo and track length more precisely.
- "Selective Section Painting" lets users edit specific track sections.
- Users can fine-tune vocals, drums, bass, and other elements.
Training data details remain limited
The source also raises the question of training data. When Google launched Lyria 3, the company said it had trained the model on materials that YouTube and Google had the right to use under their terms of service, partner agreements, and applicable law.
Google did not provide details about the training data at that time. Asked about the training data for Lyria 3.5, Google didn't immediately respond.
That leaves the public description of Lyria 3.5 focused mainly on product capabilities rather than the underlying dataset. The available facts are about the model’s output, its editing workflow, its controls, and its availability through Google Flow Music.
Why the update matters
Lyria 3.5 appears designed to make generated music less of a one-shot process. Better melodies, better lyrics, clearer vocals, and more precise tempo and duration controls all support the same basic goal: helping users move from an initial generation to a more deliberate result.
The new section-editing feature is especially important because it changes the cost of revision. If a user can adjust only the part that needs work, the creative process becomes less dependent on repeatedly generating full tracks from the beginning.
For now, the key facts are clear: Google has released Lyria 3.5, made it available through Google Flow Music, and added tools for section-level edits, tempo control, track length control, and element-specific adjustment across vocals, drums, bass, and more.