Google has unveiled Lyria 3 Pro, a major upgrade to its AI music generation model that can now create full-length tracks up to three minutes, a significant leap from the 30-second clips offered by the standard Lyria 3. Released on March 25, 2026, the new model features structural composition awareness, allowing users to specify song elements like intros, verses, choruses, and bridges. Lyria 3 Pro supports multimodal inputs including text, images, and video prompts, generating high-fidelity music complete with vocals and auto-generated lyrics. The model is available across six Google platforms including Gemini App, AI Studio, and Vertex AI, with daily generation quotas based on subscription tier.

Google has unveiled Nano Banana 2 (Gemini 3.1 Flash Image), a significant upgrade to its AI image generation model that delivers Pro-level quality at Flash speed. Launched on February 26, 2026, the new model generates images 3-5 times faster than its predecessor while maintaining approximately 95% of Pro quality. Key improvements include enhanced text rendering that fixes previous blurry text issues, advanced subject consistency for up to five characters, real-time knowledge integration, and support for resolutions from 512 pixels to 4K. The model is available for free through the Gemini app and offers 50% lower API pricing compared to Nano Banana Pro.