Google has officially released Lyria 3.5, the latest version of its music generation model, which is now available for users within the Google Flow Music platform. This update focuses on refining the core components of AI-assisted song creation, specifically targeting improvements in musicality, lyrical coherence, and vocal performance. The rollout, which began on July 29, 2026, aims to provide creators with more sophisticated tools for generating complex melodic structures.
A primary focus of the Lyria 3.5 update is the enhancement of vocal quality. According to Google, the model is now capable of producing more realistic and emotionally nuanced singing voices. These improvements extend to pronunciation, allowing the generated vocals to sound more natural and expressive than previous iterations. This development is intended to help users craft tracks that feel more authentic and less mechanical in their delivery.
The update also addresses the structural aspects of songwriting through enhanced lyric generation. Google states that Lyria 3.5 demonstrates better adherence to user prompts and a higher level of structural awareness when composing lyrics. This means the model is better equipped to maintain consistent themes and rhythmic patterns throughout a song, reducing the need for manual corrections during the creative process.
Beyond the quality of the output, Google has introduced new features designed to give users greater creative control over their compositions. Users can now more easily adjust the tempo and the overall duration of their generated tracks directly within the Flow Music interface. These adjustments are intended to provide a more flexible workflow, allowing creators to tailor the output to fit specific project requirements or stylistic preferences.
The integration of Lyria 3.5 into Google Flow Music represents a continued effort by the company to refine its generative audio capabilities. By focusing on the intersection of technical musicality and user-facing controls, the update seeks to streamline the production of richer and more complex musical arrangements. Users can access these new features immediately through the existing Flow Music platform to begin experimenting with the updated model's capabilities.