Summary
Highlights
NotebookLM now offers 'cinematic video overviews,' transforming research and documents into fully generated videos with fluid animations, detailed visuals, and proper narrative structures. This feature utilizes Gemini 3, Nano Banana Pro, and Veo3 models and is available to Google AI Ultra subscribers.
Google launched Nano Banana 2 (technically Gemini 3.1 Flash Image), offering pro-level image quality at high speed. It can maintain consistency across up to five characters and 14 objects, significantly improves text rendering within images, and supports up to 4K resolution with full aspect ratio control. This is rolling out across various Google platforms and is a free upgrade for Flow users.
Google Labs' free tool, Pomelli, now has a 'photo shoot' feature, allowing small businesses to transform phone photos of their products into professional studio-quality shots using Nano Banana and their 'business DNA'. It also includes style reference images, conversational editing, and a campaign builder that integrates with product URLs.
Lia 3, Google's music generation model, is now integrated into the Gemini app. Users can describe desired music, and it generates 30-second tracks with instrumentals, vocals, and original lyrics. It offers extensive control over genre, tempo, mood, and even allows pairing a photo with a prompt to generate a matching track. This feature is free and available in all countries where the Gemini app is accessible.
Chrome introduces 'auto browse,' an agentic capability that allows Chrome to take actions across websites on behalf of the user. This includes tasks like scheduling appointments, filling out forms, collecting documents, comparing travel options, and even identifying and purchasing items from photos. It operates within a Gemini side panel and requires user confirmation for sensitive actions. Currently in preview for Google AI Pro and Ultra subscribers in the US.
Gemini 3 Deep Think, a specialized reasoning mode for science, research, and engineering problems, has received a significant upgrade. It has achieved impressive benchmark scores on complex tests, including identifying logical flaws in research papers and assisting in cutting-edge scientific experiments. This mode is available in the Gemini app for Google AI Ultra subscribers, with early API access for researchers and engineers.