Gemini app now lets you create 3-minute songs with Lyria 3 Pro

Google launched Lyria 3 Pro, which extends generated song length to up to 3 minutes (up from the prior 30-second capability) and adds better musical-structure control (intros, verses, choruses, bridges). It is available today in the Gemini app for Google AI Plus (10 tracks/day), Pro (20 tracks/day) and Ultra (50 tracks/day), is rolling into Google Vids/Workspace and is in public preview on Vertex AI for enterprise, plus availability in AI Studio, the Gemini API, and ProducerAI. Google emphasizes safety and IP controls, stating training materials are ones it has rights to use, that the model cannot mimic artists, outputs are checked against existing content, and all generations carry SynthID watermarks.

Analysis

This product push accelerates the race from model novelty to platform utility: the immediate commercial lever is API and cloud consumption, not music chart hits. Expect incremental Google Cloud/Ads/API revenue to compound over 12–24 months as studios, gaming companies, and SaaS vendors embed on-demand music generation into pipelines — a few high-volume customers can move cloud revenue by low-single-digit percentage points which, for a large cloud vendor, is material to growth multiples.

Compute demand is the hidden supply-side winner. Large-scale, low-latency audio generation shifts workloads from research bursts to persistent inference and fine-tuning, favoring vendors with proprietary accelerators and optimized stacks; that raises the bar for smaller inference providers and increases bargaining power for infrastructure leaders over the next 6–18 months. At the same time, provenance tools (watermarking, content-ID) create a new compliance and metadata market — expect demand for detection, licensing automation, and legal services, which will monetize more slowly but stickier than one-off track generation.

Regulatory and adoption risks are asymmetric and front-loaded. Copyright litigation, regional data-usage restrictions, or a high-profile provenance failure could pause enterprise rollouts in weeks to months, compressing valuation multiples for pure-play music platforms faster than for diversified cloud/platform owners. Conversely, if developer uptake (Vertex/Studio API metrics) shows sustained monthly growth, the market will re-rate platform owners within 6–12 months, while small incumbents face structural margin erosion from commoditization.

The consensus that “AI music = immediate new ad product” understates two forces: (1) pricing compression as generative audio becomes commoditized, and (2) the winner-takes-most dynamic for distribution and billing. That implies platform exposure (distribution + billing) is underpriced relative to single-product music catalogs; position sizing should favor large-cap distribution/cloud franchises over niche music licensors until regulatory clarity arrives.

Key Decisions for Investors

• GOOGL — Overweight (12–18 months). Buy a modest call spread or add to core overweight: asymmetric upside from increased Cloud/API monetization and cross-sell versus limited downside if ad/cloud growth moderates. Target 2:1 reward:risk versus a 15–20% capital allocation to the trade leg.
• NVDA — Long (6–12 months). Buy calls or add to exposure to capture incremental inference demand from audio/creative workloads; expected >20% upside in a strong adoption scenario. Tail risk: near-term selloffs driven by macro/earnings; size accordingly (smaller theta-sensitive position).
• SPOT or SSTK — Tactical short (3–12 months). Buy puts or short equity in pure-play music/stock-audio platforms that lack diversified distribution/billing: downside if AI substitutes stock libraries. Risk: partnership deals with platform owners could offset losses; tether position size to catalyst (quarterly revenue misses).
• ADBE — Long (9–18 months). Buy calls or add small tactical exposure to capture monetization of integrated creative AI workflows (licensing, subscriptions, enterprise lock-in). Downside risk is slower conversion of enterprise users; cap position to 3–5% portfolio risk.