Google is expanding its Gemini AI with two major updates: Canvas, a structured workspace for writing and coding, and Audio Overview, a feature that generates spoken summaries of key topics.
These additions come after Google’s recent decision to transition away from Google Assistant, positioning Gemini as a more comprehensive AI assistant for productivity and research.
With these tools, Google aims to provide a more interactive AI experience, allowing users to generate, edit, and consume information in different formats. The company also plans to extend Canvas and Audio Overview to tablets, smartwatches, and other connected devices later this year.
Canvas: A Dedicated Workspace for AI-Assisted Writing and Coding
Unlike Gemini’s traditional chat-based interface, Canvas offers a structured workspace where users can interact with AI-generated content without disrupting the overall document.
The tool supports real-time code refinement, focused text editing, and structured content workflows, making it more than just a conversational AI.
Google’s approach appears to be a direct response to OpenAI’s ChatGPT Canvas, introduced in October 2024, which brought a similar feature to its AI ecosystem.
However, Google’s implementation integrates Gemini’s AI-driven suggestions with Google Search and Workspace tools, offering a more interconnected editing experience.
Canvas reduces the time spent on document revisions and code debugging substantially, thanks to its ability to isolate and refine AI-generated content without manual rewrites.
Audio Overview: AI-Generated Summaries for Hands-Free Consumption
Alongside Canvas, Google is launching Audio Overview, a feature that delivers spoken summaries of key topics. Designed for users who prefer auditory learning or need hands-free access to information, this tool enhances how users interact with Gemini.
Google has improved speech synthesis in Audio Overview to make it more engaging and structured, unlike previous AI-generated narrations from Google Assistant. Audio Overview provides context-aware summaries rather than reading content verbatim, making it easier to grasp essential information quickly. In its initial rollout, Audio Overview supports only English, but additional language support is expected in the coming months.
Google Shifts to Gemini as its Primary AI Platform
The updates are part of Google’s broader strategy to phase out Google Assistant and transition entirely to Gemini. As Google confirmed, Assistant will no longer be accessible on most Android devices as users are upgraded to Gemini.
In addition to Canvas and Audio Overview, Google has recently made advanced features like Deep Research and Custom Gems free for all users. These tools, previously available only to premium subscribers, allow for multi-step research tasks and AI-powered personalization.
This move positions Google Gemini as a direct competitor to OpenAI’s Deep Research feature and Perplexity AI with its own Deep research implementation.
Additionally, Google has recently introduced a personalization feature that leverages users’ search history to tailor AI responses. While this aims to improve relevance, it has also raised privacy concerns regarding data usage and AI-driven personalization.
Gemini vs. OpenAI and Anthropic: The AI Assistant Race
Google isn’t alone in redefining AI-powered workspaces. OpenAI’s ChatGPT Canvas recently introduced live HTML and React code previews, making it a strong competitor to Gemini’s Canvas.
Anthropic’s Claude provides an Artifacts workspace that allows users to generate, edit, and organize AI-generated content in a structured environment.
These developments indicate that AI-powered assistants are evolving beyond simple chatbots, becoming integrated tools for productivity and research.
Looking ahead, Google plans to expand Canvas and Audio Overview beyond smartphones, integrating them into Wear OS smartwatches and Google Home devices.
Both features are expected to play a key role in Google’s push toward a fully AI-driven ecosystem, where Gemini adapts dynamically to users’ needs across different devices.
