Gemini adds music, reasoning, and images
Google's February Gemini Drop adds music creation with `Lyria 3`, advanced reasoning with `Gemini 3.1`, image generation with `Nano Banana 2`, and video templates. It also offers direct links to scientific papers and reserves `Deep Think` for Google AI Ultra subscribers.
The Gemini app is adding new tools for creating music, generating images, producing videos, and solving complex problems. The updates are part of Google's February Gemini Drop and already include beta features alongside options limited to certain subscribers.
30-second music from text or images
Lyria 3, which Google describes as its most advanced music model, can create tracks of up to 30 seconds from a written description or an image. The feature is available in beta.
For example, you can upload a travel photo and ask for a soundtrack that matches its atmosphere, or describe the style you want to generate a custom piece for a video or presentation.
More capacity for difficult problems
Gemini 3.1 includes improvements for solving complex problems. The update includes 3.1 Pro, designed for demanding tasks, and Deep Think, a specialized reasoning mode for modern science and engineering.
Deep Think is available to Google AI Ultra subscribers. It is not presented as a general feature for all users, but as a specific option for people who need to work on more difficult technical problems.
More accurate text in images
Google is also adding Nano Banana 2, an image generation model designed to offer greater fidelity and speed. One of its main features is the ability to add text in any language with greater accuracy compared with the requested result.
This can be useful for creating posters, covers, social media posts, or visual materials without having to manually correct every word that appears in the image.
Templates to get started with a video
Veo Templates offers a gallery of styles that you can use as a starting point for creating videos. The process involves choosing a template, using it as a reference, and modifying it with your own details.
The idea is to reduce the initial work: instead of starting from a blank screen, you can begin with a defined visual concept and adapt it to the result you want.
Direct scientific citations
Gemini also makes it easier to find academic sources. It can now provide direct links to verified scientific papers, so you can find the original document without having to review search result pages one by one.
The feature does not remove the need to check the content and context of each study, but it does make it easier to reach quality sources when you are researching a topic.
The most important change in this update is its range of uses: Gemini does not just answer questions, it is also becoming a tool for creating, researching, and automating parts of your work. It is worth watching which features reach all users, which remain in beta, and which require a subscription such as Google AI Ultra.