What you need to know
- Google’s Gemini Drop for November recaps all the features and changes to hit the Gemini app this month.
- Google swapped three models, debuting Gemini 3, Veo 3.1 with Ingredients to Video, and Nano Banana Pro.
- The app also added a redesigned interface with a My Stuff tab dedicated to generative AI creations.
Google was hard at work upgrading the Gemini app in November, giving users a new large language model (LLM), image-generation model, and video-generation model. These updates rolled out gradually throughout the month, and are recapped in Google’s latest Gemini Drop for November. If you haven’t followed every singe announcement, this is a great way to figure out the new Gemini app you need to try.
The big change is the addition of Gemini 3, the company’s latest and greatest LLM. Google calls it the “most intelligent model” it has ever made, with deep understanding and stellar vibe coding abilities. The update first debuted on the Gemini web client, and now, it’s available in the Gemini app on iOS and Android.
To use it, tap the current model listed at the bottom of the Gemini app home screen — usually 2.5 Flash — to open the model selector. Then, tap Thinking to switch to the preview Gemini 3 Pro model, the first one to release as part of the next-generation Gemini 3 series. The release corresponds with a new Gemini mobile app view, which adds a My Stuff page for images, videos, and Canvas content.
Google also changed out the Gemini app’s image- and video-generation models this month. Nano Banana Pro (Gemini 3 Pro Image) replaces Nano Banana (Gemini 2.5 Flash Image) in the Gemini app as the default image generator, at least to start. Free users get a few free generations before the model reverts back to Nano Banana, while Google AI Plus, Pro, and Ultra subscribers get higher rate limits.
For videos, Veo 3.1 is the default video-generation model, and it now supports the Ingredients to Video feature first made available in Flow. It allows users to upload three images that will guide a video prompt, and it aims to shorten the length of prompts needed to generate a requested video. Between the addition of Nano Banana Pro and Veo 3.1 Ingredients to Video, it’s clear the new My Stuff section of the Gemini app is part of a content-generation push.
For Google AI Ultra subscribers, the cornerstone feature added in November is the Gemini Agent. It can take actions on your behalf in the Google ecosystem and beyond, all while prioritizing user control. Gemini Agent is based on Project Mariner, the browser use tool that began testing in 2024.
Gemini Live also added “more adaptive and expressive” conversations this month, allowing users to customize the speed and tone of the AI assistant’s responses.
What’s next for Gemini?
Google had a busy month shipping Gemini features in November, and the compilation of them in a November Gemini Drop suggests we’ve already seen the best additions this month has to offer. There’s plenty to look forward too, though.
The company is currently safety testing Gemini 3 Deep Think, an even smarter AI model that will come to Google AI Ultra subscribers in the coming weeks. Google also confirmed that more Gemini 3 models will be released “soon” to join Gemini 3 Pro in preview. We’re likely to see a lightweight model, like Gemini 3 Flash or something similar, to replace or join Gemini 2.5 Flash for basic queries.
Google could also expand limited features, like Gemini Agent, to more users in the future. We’ll have to wait until the December Gemini Drop to find out.
