Create Google Docs With Your Voice

 

“Microsoft Word brought word processing to the masses.

 

Then Google Docs pioneered real-time collaboration with our fellow humans.

 

Now Google wants to usher in a new revolution: the AI collab -- no typing necessary. You compose with your voice, and your document instantly takes shape. It isn't for aspiring Shakespeares, but it is for anyone who wants to spin up a quick reference guide, checklist or set of instructions.

 

Google has long excelled at converting speech to text. Its Pixel phone routinely out-transcribes the iPhone. With generative AI, machines no longer just transcribe speech, they understand what we're saying.

 

Google is leaning in to more voice-powered products. Gmail Live, announced at this week's Google I/O conference, allows you to ask your inbox questions aloud, like what's your flight reservation number or the door code to your Airbnb.

 

I got early access to another product Google announced: Docs Live. It combines AI conversation mode with word processing to get you started on a first draft. Instead of being faced with a blank page, you just start talking.

 

Docs Live operates in two stages: First comes your rambling, followed by a chat with the Gemini AI about your high-level goals. Then you work through the actual draft, asking the AI to do things like rearrange sections or adjust the tone.

 

What better test subject than this very column? I tapped Docs' new "Create with voice" option and spewed a stream of verbal garbage, full of "ums," sentence fragments and idea pivots.

 

The bot blundered early. I asked for an outline, but the overeager AI composed a full draft instead. I forced a do-over. (And for the record: My words here are 100% human-generated.)

 

As Gemini and I went back and forth on the structure, I started to see the light. The AI sifted through my Google Drive for relevant interview transcripts and searched the web for facts at my bidding.

 

After about five minutes of chitchat, we landed on a satisfying framework. I disengaged this autopilot by saying "Generate the document," then continued to write and edit myself, as usual.

 

Soon, I was speaking drafts into existence left and right. Within an hour, I conjured a performance review, a project postmortem and a meal plan engineered to satisfy a picky toddler and use up every veggie in my fridge.

 

I hit some other bumps in the road as I tinkered.

 

I requested a checklist that looked like a mess even after Gemini proudly proclaimed, "OK, I've taken care of that for you." I pleaded with the AI to fix formatting that would take just a few of my own keystrokes.

 

Google Docs already has two AI buttons built in -- a chat side panel and a bottom bar. Does it need another window into Gemini?

 

Frank Tisellano, a Docs product manager, says yes, especially for people who think and speak faster than they type: "We're trying to remove this bottleneck to prompting and getting your ideas out there."

 

My prerelease build has some kinks that will be ironed out before launch, he added.

 

Voice-based apps such as Wispr Flow are gaining traction.

 

 Google's biggest advantage here may be that it already has access to emails, files and other documents -- things you need to get started, and things you might not want to copy over to other platforms.

 

Competing AI models "might be powerful and capable, but unless they're grounded in what's important to you, who are you working with and all of that, then it ends up being generic," says Yulie Kwon Kim, vice president of product at Google Workspace.

 

I agree it's easier to talk out loud than write thoughts down. I often get blank page paralysis when I open a new doc. My editor and I often riff on story ideas before landing on a specific angle. If he's busy when I need inspiration, I guess I could turn to AI instead. If anything, just thinking out loud helps.

 

This initial test run left me dreaming of a better version. The editing workflow is a little clunky, and I wanted to highlight sections to discuss and track the AI's changes -- more granular control. And right now, Docs Live handles everything until I take over. Ideally, I would like to type whatever I want but then call on it whenever I need it.

 

Tisellano says that change-tracking will be live at launch. In the coming months, the team will begin developing collaborative features like those on my wishlist.

 

Subscribers to Google's paid Pro and Ultra AI plans will get access to Docs Live for iOS and Android in English first.

 

The writing is pretty generic. I wasn't expecting Joan Didion-level prose, but I wanted the words to sound like they came from me. Maybe it's good that I'm still needed, though I could see Docs Live being particularly useful for non-native speakers or folks who just aren't great scribes. Either way, it's a helpful tool when you just need to get thoughts on the page.” [1]

 

1. Create Google Docs With Your Voice. Nguyen, Nicole.  Wall Street Journal, Eastern edition; New York, N.Y.. 20 May 2026: A11.