According to Google Gemini on X (Twitter), Gemini users on AI Plus, Pro, and Ultra tiers can now generate music tracks up to 3 minutes using Lyria 3 Pro, including lyric-enabled, high‑fidelity compositions from photos or text prompts (source: Google Gemini). As reported by Google Gemini, the update expands prior clip limits and positions Lyria 3 Pro as a turnkey generative audio model inside Gemini for consumer and creator workflows (source: Google Gemini). According to Google Gemini, this enables music creators, marketers, and short‑form video producers to rapidly prototype soundtrack ideas, branded jingles, and social ads, lowering production costs and time to publish (source: Google Gemini). As noted by Google Gemini, access is limited to paid Gemini plans, signaling a monetization path for long‑form generative audio and potential upsell opportunities to Pro and Ultra subscribers (source: Google Gemini).

According to Google Gemini on X (@GeminiApp), Lyria 3 is designed for rapid, fun music generation and easy sharing inside Gemini, while Lyria 3 Pro offers enhanced musical flow and deeper customization for more polished outputs, available to Google AI Plus, Pro, and Ultra users. As reported by Google Gemini, creators seeking quick ideation can use Lyria 3 to sketch hooks and short tracks, whereas producers needing control over arrangement, dynamics, and style coherence should opt for Lyria 3 Pro to elevate production quality. According to the same source, the tiered access model signals a freemium upsell path for AI music tooling, creating business opportunities for subscription bundles, creator workflows, and licensing-ready outputs within Google’s ecosystem.

According to Ethan Mollick on X, Google’s new Lyria 3 Pro music AI can transform text and poetry prompts—such as Rilke’s First Elegy—into stylistically targeted songs like a “1990s boy band” rendition, demonstrating high-fidelity style transfer and catchy vocal hooks (as reported by Ethan Mollick). According to Mollick, the system reliably follows creative direction, implying strong prompt adherence and controllability that could streamline songwriting, ad jingle production, and rapid prototyping for labels and creators (as reported by Ethan Mollick). According to Mollick, real-time iteration makes it feasible to produce multiple branded cuts quickly, suggesting opportunities for subscription tools, creator platforms, and enterprise media pipelines that need fast, consistent music variants (as reported by Ethan Mollick).