Lyria 3

Overview

Lyria 3 is a generative music and audio model from Google/Google DeepMind that can turn text prompts and, in some cases, images or photos into original music. Based on newsletter coverage, it first appeared as a beta capability in the Gemini app and later became available via API, expanding it from a consumer-facing creative feature into a developer-accessible tool.

For AI Product Managers, Lyria 3 matters because it represents the shift of multimodal generative media from app features to platform primitives. It enables teams to prototype music generation experiences, soundtrack creation, audio personalization, and branded creative workflows directly inside products. Its rollout also highlights important product questions around access tiers, clip length, watermarking, safety, and the transition from in-app experimentation to API-based platform adoption.

Key Developments

• 2026-02-19 — Google DeepMind rolled out Lyria 3 in global beta via the Gemini app for custom audio generation. Newsletter coverage noted that outputs include SynthID, Google’s invisible watermark for AI-generated content.
• 2026-02-21 — Google AI highlighted Lyria 3 as a model that can turn photos or text into dynamic music, alongside other launches such as Gemini 3.1 Pro. This positioned Lyria 3 as part of Google’s broader multimodal AI product stack.
• 2026-04-10 — Josh Woodward announced full access to Lyria 3 in the Gemini app, with up to 5 free full-length (~3 minute) tracks per day, unlimited 30-second clips, and paid upgrades for additional music, image, and video generation.
• 2026-04-11 — DeepLearning.AI described Lyria 3 as an AI music generator that transforms text prompts or images into original 30-second songs, reinforcing the core use case and accessible creative workflow.
• 2026-05-26 — Logan Kilpatrick announced that Lyria 3 is now available via API, allowing developers to build applications directly on top of the model.

Relevance to AI PMs

• Prototype new media experiences quickly: Lyria 3 gives PMs a way to test product concepts such as soundtrack generation, creator tools, music personalization, short-form audio for marketing, or music layers for video/image workflows without building generation infrastructure from scratch.
• Design monetization and usage packaging: Newsletter mentions point to clear packaging decisions—free daily full-length tracks, unlimited short clips, and paid upgrades. PMs can use this as a benchmark for thinking through quotas, premium tiers, trial design, and cost-controlled user engagement loops.
• Plan trust, safety, and provenance features early: The mention of SynthID watermarking is especially relevant for AI PMs building media products. It signals the need to incorporate provenance, disclosure, moderation, and rights-sensitive guardrails into audio generation workflows from day one.

Related

• Google / Google AI / Google DeepMind — The organizations behind Lyria 3’s launch, positioning, and rollout across app and API surfaces.
• Gemini app / GeminiApp — The initial product surface where Lyria 3 was introduced in beta and later expanded for broader user access.
• API — Marks the transition of Lyria 3 from a feature inside Google products to a developer platform capability.
• SynthID — Google’s invisible watermarking system applied to Lyria 3 outputs, relevant for provenance and responsible AI deployment.
• Josh Woodward — Mentioned in connection with expanded Gemini app access and consumer packaging for Lyria 3.
• Logan Kilpatrick — Announced API availability, signaling developer-facing adoption.
• DeepLearning.AI — Covered the launch and described the model’s text/image-to-music generation capability.