Google has added new AI-powered creation tools to its AI Pro subscription package: Google Pics, which provides advanced image generation, and agentic video understanding, which offers in-depth analysis of video content.
Google Pics is the latest advance in Google's image generation tools. Powered by its Nano Banana image generation model, Pics provides next-level generation capacity including advanced editing and customization tools.
As shown in the demo video, Pics offers a range of advanced options including language translation and the capacity to add and edit elements within a generated image. That could make creating promotional material for a range of purposes much easier — and enable a higher level of customization to help avoid the sometimes tacky AI-generated look.
Google's agentic video understanding tool processes video inputs and provides information on what was presented and who presented it. It can also offer specific details on elements shown within the clip.
The advanced video processing tool makes it easier to extract insights and information from video clips, with improved response accuracy and speed as well as lower token usage. That opens up a range of use cases for building contextual understanding and learning from video presentations.
Google said the agentic video understanding model is rolling out to all users in the Gemini app across Flash and Flash-Lite models soon.
Pics will also soon power YouTube's "Ask YouTube" feature on the video watch page, "leveraging Gemini to deliver higher-quality answers grounded in the visuals," the company said.
The two tools work in opposite directions: one makes assets, one reads them.
On production, the editing and customization matter most. The practical blocker with AI imagery has been that you get a decent image but can't bend it to brand specifications. Being able to edit elements inside a generated image narrows that gap.
On analysis, the tool can be pointed at competitor videos and creator content at volume. Extracting what was shown and who said it removes work people currently do by watching and summarizing manually.
What doesn't change is that generated assets alone aren't enough. The principle that video of real experts and real product becomes the primary source in AI search holds regardless of tooling. And since images have a new job in AI search, it's worth checking how the images you generate actually read to a machine.
An image generation tool in the Google AI Pro subscription, powered by the Nano Banana model. It includes advanced editing and customization, language translation options, and the ability to add and edit elements within a generated image.
It processes video inputs and reports what was presented and who presented it, plus specific details on elements in the clip — with better accuracy and speed and lower token usage.
The agentic video understanding model is rolling out to all users in the Gemini app across Flash and Flash-Lite soon. Pics will also power YouTube's 'Ask YouTube' feature shortly.
To apply what you just read to your own site, start with a free audit of where things are now.
A strategist replies within 24 hours on business days.
TikTok’s Road to Black Friday series shows how top TikTok Shop sellers prep for BFCM: Princess Polly’s 6,000 creator videos a month, Dossier’s sampling engine and SharkNinja’s LIVE strategy.
Meta shared how small businesses use its Muse AI agents: analyzing ad performance, building campaigns through conversation and automating workflows with Shopify, Canva and HighLevel. Prompts included.
Reddit outlined its U.S. midterm election plans, including the Up the Vote hub, AI detection that cut spam exposure 20% and removes nearly 2 million fake upvotes daily, and AMA rules for political ads.