Google updates the Gemini app with real-time AI video, Deep Research mode, and other new features to boost productivity and content creation tools
During Google I/O 2025, Google disclosed numerous enhancements to the Gemini AI chatbot app, such as deeper integrations with Google’s array of products, more widely available multimodal AI features, and updated AI models.
Starting on Tuesday, Google will extend the camera and screen-sharing capabilities of Gemini Live to all iOS and Android users. The feature, which Project Astra fuels, enables users to engage in near-real-time verbal conversations with Gemini while simultaneously streaming video from their smartphone’s camera or screen to the AI model.
For instance, while traversing a novel city, users could aim their smartphone at a structure and inquire about its architecture or history with Gemini Live. Responses will be provided promptly.
Google has also announced that Gemini Live will begin to integrate more thoroughly with its other applications in the coming weeks. Gemini Live will soon be capable of providing coordinates from Google Maps, creating events in Google Calendar, and creating to-do lists with Google Tasks, according to the company.
The numerous updates to Google’s Gemini are part of the company’s endeavors to compete with OpenAI’s ChatGPT, Apple’s Siri, and other digital assistant providers. The emergence of AI chatbots has provided users with a novel method of interacting with the internet and their devices.
This has significantly pressured numerous large technology companies, such as Google Search and Google Assistant. During I/O 2025, Google disclosed that Gemini had 400 million active users monthly. The company is confident that these upgrades will contribute to expanding its user base.
Google has launched two new AI subscriptions: Google AI Pro, which is a rebrand of its $20-per-month Gemini Advanced plan, and Google AI Ultra, a $250-per-month plan that directly competes with ChatGPT Pro. The Ultra plan provides users exclusive access to specific features, early access to new AI models, and very high rate limits.
Google announced on Tuesday that Gemini will be accessible in the Chrome browser for U.S. subscribers to Pro and Ultra who have selected English as their language in Chrome. The integration’s objective is to enable users to request that Gemini summarize information or respond to inquiries regarding the content displayed on their screen.
Google has announced that it enhances Deep Research, Gemini’s AI agent that produces comprehensive research reports, enabling users to upload their private PDFs and images. Deep Research will cross-reference these private PDFs with public data to generate more personalized reports. According to the company, Users will soon be able to integrate Drive and Gmail with Deep Research directly.
An updated AI image model, Imagen 4, is being provided to free Gemini users. According to Google, this model produces superior text outputs.
In addition, subscribers to the company’s new $250-per-month Google AI Ultra plan will have access to Google’s most recent AI video model, Veo 3, which also generates sound corresponding to video sequences through native audio generation.
Google is also upgrading the default model in Gemini to Gemini 2.5 Flash, which the company claims will provide superior-quality responses with reduced latency.
Google has announced that Gemini will now develop personalized exams that concentrate on the areas users find challenging to accommodate the increasing number of students who utilize AI chatbots.
Gemini will assist in developing supplementary quizzes, and action plans to enhance the areas where users answer queries incorrectly.
