Google just made the Gemini app a whole lot more useful. Today’s update brings three major additions: Audio Overviews, a NotebookLM-inspired feature that turns documents into AI-generated podcast-style discussions; a new Canvas mode for real-time coding and document editing; and an expansion of Deep Research queries. These updates will make Gemini an even more powerful tool for both casual users and professionals.
Key Points:
- Canvas mode lets users write, edit, and preview code in real time, with live feedback from Gemini.
- Audio Overviews generate AI-powered discussions summarizing documents and research reports.
- Deep Research now supports more free queries, expanding access to detailed AI-generated insights.
Google is continuing its push to make the Gemini app a must-have AI assistant. Today’s update introduces Canvas, an interactive workspace where users can write and refine documents or code with Gemini’s assistance. The feature supports real-time editing, code previews for HTML and React, and instant content adjustments—making it an all-in-one space for creating, editing, and sharing work. Developers can use it to prototype web apps, debug code, and generate previews, while writers can fine-tune tone and structure with AI-driven feedback.
Welcome to the Gemini app, Audio Overviews 🎧
— Google Gemini App (@GeminiApp) March 18, 2025
We've seen incredible excitement around Audio Overviews in @NotebookLM, which helps people make sense of complex information. Today, we're making Audio Overviews available in the Gemini app, too.
Transform your documents, slides,… pic.twitter.com/xsWslnn05V
Another highlight is Audio Overviews, a feature lifted from Google’s NotebookLM tool. It turns documents, research reports, and even slide presentations into dynamic AI-generated discussions between two virtual hosts. Instead of reading through lengthy materials, users can now listen to engaging, podcast-style conversations that summarize key insights.
Lastly, Deep Research, Gemini’s AI-driven research tool, is expanding its availability, allowing more users to access detailed AI-generated insights for free.
These enhancements reflect Google's ongoing efforts to expand Gemini's capabilities beyond simple text generation and into more specialized productivity and creative domains, areas where competitors like Microsoft's Copilot, ChatGPT, and Anthropic's Claude have been building their offerings.
Canvas is rolling out globally today for both free Gemini users and Gemini Advanced subscribers in all languages where the app is available. Audio Overviews are initially launching in English only, with additional languages planned for future updates.