Google brings conversational voice AI to Gmail, Docs and Keep with its 'Live' features

Google has begun rolling out a set of conversational artificial intelligence features within Gmail, Docs and Keep that allow users to interact with these applications through natural language and voice dictation, rather than relying solely on the keyboard or traditional search boxes. The company is presenting them under the names Gmail Live, Docs Live and Keep Live, and had already previewed them in May during its Google I/O conference.

In Gmail, the new feature allows users to put questions directly to a Gemini-based assistant about the contents of their inbox, instead of typing search terms. While conversing with the assistant, the user sees a live transcript of the interaction, providing transparency about what the system is interpreting and answering.

In Google Docs, the logic is similar but geared toward writing: the user verbally describes what they want to write and the AI generates a first draft. To do so, the assistant can draw on information from Gmail, Drive, chat and the web, combining different internal and external sources to build the initial text, which the user can then edit.

Keep Live, for its part, turns the notes application into a kind of voice notepad: it allows users to take notes without typing and, according to the example cited in the article itself, it can organize lists or recipes from a spoken request, such as asking it to list the ingredients needed to prepare a shakshouka.

As for availability, Gmail Live is reaching subscribers of Google AI Plus, Pro and Ultra, while Docs Live and Keep Live are reserved, for now, for the Pro and Ultra plans. All of these features are launching initially in English only and on the iOS and Android versions; Google has indicated it will extend them soon to its Workspace Business enterprise customers.

The launch is part of a broader trend within Google toward Gemini-powered voice features. The article notes that the company had already added cross-app dictation to its Mac app in April, that Pixel 11 phones included a dictation tool called Rambler last month capable of removing filler words from speech, and that in August it introduced the Gemini 3.5 Transcribe model for speech-to-text tasks. Taken together, these pieces suggest that Google is incrementally building a voice interaction layer that spans several of its products, from hardware to Workspace productivity applications.

The original article is a TechCrunch 'In Brief' item, so the available information is limited: it does not detail adoption metrics, exact pricing for the AI Plus, Pro or Ultra plans, or direct comparisons with voice tools or assistants from other companies. Any assessment of these features' competitive impact should be taken with caution, since the text merely describes the launch and its immediate availability.