openaienmodel: gpt-5-mini-2025-08-07
Find out what’s new in the Gemini app in February's Gemini Drop.
Key Points
- Lyria 3 beta: 30s music from text/images
- Gemini 3.1 + Deep Think for complex reasoning
- Nano Banana 2: multilingual text in images
Summary
Gemini Drop (published 2026-02-27) adds multimodal creation and stronger reasoning to the Gemini app. Highlights include Lyria 3 (beta) for short music generation from text or images, Gemini 3.1 with a Pro tier and a Deep Think mode for advanced scientific/engineering problems, Nano Banana 2 for higher-fidelity images with accurate multilingual text, Veo Templates for rapid video remixing, and verified scientific citations that link directly to papers. See the Gemini Drops Hub for examples and rollout details.
Key Points
- Lyria 3 (beta): generate 30-second soundtracks from text or images — fast audio prototyping for prototypes and demos.
- Gemini 3.1: improved reasoning and problem solving; 3.1 Pro targets demanding workflows; Deep Think provides specialized scientific/engineering reasoning (requires Google AI Ultra).
- Nano Banana 2: higher-fidelity image generation that renders text accurately in any language and improves generation speed.
- Veo Templates: gallery of editable video templates to accelerate polished video creation and iteration.
- Verified scientific citations: Gemini now provides direct links to research papers to validate results and support reproducibility.
- Where to find it: details, examples, and availability notes on the Gemini Drops Hub.