You still type emails like it’s 2015, and honestly, same. But the typing era is getting a serious challenger: Google AI voice features are rolling out to subscribers right now, and they let you talk to Gmail, Docs, and Keep instead of typing into them. These aren’t the old “dictate into a text box” tricks. These are full conversations with your apps.
Google calls them Live features: Gmail Live, Docs Live, and Keep Live. Here’s what each one does, who gets them, and whether they’re worth your attention.
What are Google AI voice features?
At its I/O conference in May, Google previewed a big idea: instead of typing in a search box or a blank document, you talk to a Gemini-powered assistant that lives inside your apps. In September, those features started actually shipping.
They’re conversational. You don’t bark single commands like “compose email to John.” You have a back-and-forth: ask for something, change your mind, add a detail, and the assistant follows along. That’s the part that’s actually new. Google’s been clear that the feature understands when you change your mind and ask for a detail to be modified in the same conversation turn.
Three apps got the treatment, and they work differently.
Gmail Live: talk to your inbox
Gmail Live replaces the search bar with questions. Instead of typing “flight confirmation from United” and hoping, you can ask things like “what’s the price I agreed to pay for that apartment?” or “when is my car’s inspection due?”
The assistant reads your actual inbox to answer, and you see a live transcript while you chat with it. It’s basically a conversation with someone who has read every email you’ve ever received. For many people, Gmail Live is the most useful of the Google AI voice features to start with.
For the people who live in their inbox, this is the killer feature of the batch. Hunting for one email in ten years of archives is the kind of task that eats 20 minutes and makes you question your life choices. Asking a question and getting an answer is the fix.
Docs Live: dictate your drafts
Docs Live is the feature that feels like magic the first time you use it.
You describe what you want to write, and the assistant creates a first draft. Say “write a welcome email to new subscribers including the schedule for the launch week,” and it produces something you can actually edit instead of a blank page.
The useful part is context. The assistant can pull information from your Gmail, your Drive, your chat history, and the web to build the draft. In Google’s demo, someone dictated a document that pulled their résumé details from Drive, added event logistics from an email on top, and even worked in a couple of anecdotes. Everything in one conversation.
This is a serious time-saver for anyone who writes for work. Not because the first draft is perfect (it won’t be), but because the blank page problem disappears. Editing something is faster than creating it from nothing, and that’s the whole game.
Keep Live: notes without typing
Keep Live turns Google Keep into a hands-free scratchpad.
You dump thoughts into it with your voice, and the AI turns that rambling monologue into a structured note or a list. Tell it something like “list the ingredients to make a shakshouka,” and it organizes the list for you instead of transcribing your stream of consciousness verbatim.
This is the low-key winner for one simple reason: most note apps die because typing notes is friction. You think of something while cooking or driving, and by the time you’d type it, the moment’s gone. Voice removes that friction entirely.
Who gets these features
Here’s where Google draws the lines:
- Gmail Live is available for Google AI Plus, Pro, and Ultra customers
- Docs Live and Keep Live need Google AI Pro or Ultra
The features are in English, on both iOS and Android. Google says it will roll them out to Workplace Business customers soon, so if you use Google Workspace for your company, they’re probably coming to your work account too.
If you’re not a subscriber at all, you’re not missing out yet. But Google’s pattern is clear: AI voice features started on the Mac app back in April, and the company keeps pushing the same ideas into more places. We covered Gemini’s voice commands on Mac when those launched, and this is the same wave, further along.
How to try them
The Google AI voice features are rolling out to subscriber accounts over the coming days, so keep an eye on your apps.
A few tips from how these tools actually behave:
Start with a real task, not a test. Ask Gmail Live something you genuinely need to know from an old email. It’s more motivating than poking it with fake questions.
Use the conversation, don’t reset it. The whole point is that you can say “actually, change that to Tuesday” and it follows. Lean into that.
Treat drafts as starting points. Docs Live will produce something usable but not perfect. Plan to edit. That’s still way faster than writing from zero.
Combine it with your other Gemini tools. Google’s been stacking these features all year, like letting Gemini Spark manage your Google Photos library, and our guide to Gemini Gems shows how to build reusable assistants that pair nicely with the Live features.
You’ve seen this before, sort of
Voice-to-structured-notes isn’t a new idea. Apps like Voicenotes and AudioPen have been turning rambling voice memos into clean notes for a few years, and dictation tools like Wispr Flow, Monologue, and Aqua built the same trick into their typing products, as TechCrunch noted when Google previewed all this in May.
So why pay attention to Google’s version? Two reasons. It lives inside the apps where your work actually happens, instead of a separate notes app you’ll forget to open. And it has Gemini’s context powers behind it, which means it can pull from your inbox and your Drive while you talk. The standalone apps are great for personal brainstorming. Google’s version is aimed at the workday itself, as the I/O preview made clear.
The bigger picture
These Google AI voice features sit inside a much bigger voice push. Google added cross-app dictation to its Mac app in April. The Pixel 11 phones got a Gemini dictation tool called Rambler in August, which cleans filler words out of your speech. And in the same week these Live features launched, Google released its Gemini 3.5 Transcribe model for speech-to-text, as TechCrunch reported.
Microsoft is racing the other direction with its own AI transcription model, and we covered that price war separately. The short version: voice and speech are the battleground of the second half of 2026, and Google’s strategy is to make talking to your computer feel boring and normal.
The takeaway
Google AI voice features are the first real attempt to make conversation the default way you interact with your work apps. Gmail Live answers inbox questions, Docs Live drafts whole documents from a description, and Keep Live kills the friction that kills note taking.
If you subscribe to Google AI Pro or Ultra, start playing with these this week. The features are new and a little rough, but the direction is right: talking to your apps should be as normal as talking to your phone.
Your keyboard will survive. It just won’t be the only way in anymore.
