Google Lyria 3 Pro Preview is a generative audio model from Google, producing audio output from text and image inputs. Like its Clip sibling, it is a preview-stage specialised model outside the scope of standard language model benchmarking.

For digital marketing and business language tasks, this model has no direct applicability. Its relevance is limited to audio content generation workflows, which remain a niche use case for most businesses.

Teams with specific audio generation needs may wish to monitor its development as it moves beyond preview. For all standard business AI use cases, general-purpose language models remain the appropriate choice.

Assessed July 10, 2026

Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?