Talk to Your Inbox: Gmail Live, Docs Live, and Keep Live Put Gemini Audio Inside Workspace

Google has quietly shipped one of the most practical AI upgrades to hit everyday productivity software this year. Following a preview at Google I/O in May, the company has officially launched Gmail Live, Docs Live, and Keep Live — three voice-activated, conversational AI features powered by the new Gemini Audio models — across its core Workspace apps. The rollout started this week, and it signals a decisive shift in how Google thinks people should interact with their productivity tools: not by typing queries into search boxes, but by simply talking to their apps.

The announcement, posted on the official Google Workspace blog by Yulie Kwon Kim, VP of Product for Google Workspace, describes the trio of features as a way to “tackle daily tasks using just your voice” — whether that’s searching an inbox, organizing thoughts, or brainstorming a document from scratch.

Gmail Live: conversational search for your inbox

The flagship of the three is Gmail Live. Instead of constructing keyword queries and digging through endless threads, users can now ask natural questions like “What’s my flight’s gate number?” or “What’s going on at my kid’s school this week?” — and Gemini will scan the inbox and answer directly. A live transcript is displayed while chatting with the assistant, so you can see what the model heard and correct it on the fly.

This is the kind of feature that makes AI tangible for people who don’t follow AI news. Flight gates, school announcements, order confirmations, appointment reminders — the details buried in modern inboxes are exactly the thing keyword search handles poorly and conversational search handles well. Google is betting that the inbox, the most universally used productivity surface it owns, is the right place to introduce mainstream users to agentic voice interaction.

Docs Live: a hands-free thought partner

Docs Live takes a different angle: writing. Rather than staring at a blank page, users engage in real-time conversation with the assistant to build structured, context-aware documents. Docs Live acts as a co-writer that organizes your thoughts, structures the document, and — with explicit permission — pulls relevant details from Gmail, Drive, Chat, and the broader web to enrich the draft.

The “with your permission” qualifier is worth noting. Docs Live crosses a line that earlier writing assistants didn’t: it can read across your entire Google ecosystem to ground what it writes. That’s enormously powerful for drafting briefings, trip plans, or project summaries — and it’s also the kind of cross-context access that privacy observers will watch closely. Google has framed the permission model as a feature, but the implementation details of exactly when and how consent is requested will determine whether users trust it.

Keep Live: brain dumps, structured automatically

The smallest of the three might be the most delightful. Keep Live turns spoken streams of consciousness into organized, actionable notes. Ramble about your week, your shopping list, or a project idea, and Keep processes the audio in the background, returning structured lists and notes — no typing required. TechCrunch’s coverage highlighted a concrete example: ask it to list ingredients for shakshouka, and it assembles the recipe list for you.

Availability and pricing tiers

The tiering tells you where Google sees the monetization opportunity:

• Gmail Live — available to Google AI Plus, Pro, and Ultra subscribers
• Docs Live and Keep Live — restricted to Pro and Ultra tiers
• All three features are English-only at launch, on iOS and Android
• Workspace Business customers get access “soon,” per Google

The staggered tiers are a classic freemium ladder: inbox search is the broad hook (even the cheaper Plus tier gets it), while the more computationally demanding drafting and note-structuring features stay behind the premium paywall. Business availability “coming soon” suggests Google is still working through enterprise compliance and admin controls — a pattern seen with nearly every consumer-first Gemini feature that later crosses to Workspace.

Context: voice is becoming Google’s default interface

The Live trio doesn’t exist in isolation. Google has been systematically seeding voice AI across its product line all year: cross-app dictation arrived in the Mac Gemini app in April; the Pixel 11 lineup shipped with Rambler, a Gemini-powered dictation tool that strips filler words from your speech, last month; and August brought the Gemini 3.5 Transcribe model for speech-to-text workloads. The Workspace launch essentially completes the picture — Gemini Audio is now the connective tissue between Google’s phone, desktop, and productivity surfaces.

It also lands exactly one day after Google Assistant’s mobile shutdown completed its final phase, marking the end of the largest forced AI migration in consumer software history. The message is unambiguous: the old generation of command-based voice assistants is dead, and conversational, context-aware AI that can actually act inside your apps is the replacement.

What it means

Three things stand out. First, this is agentic AI packaged for people who never asked for “agents” — you speak, the AI searches, reads, and drafts across your own data. That’s the agent pitch without the jargon. Second, the permission-gated cross-app context access in Docs Live is a template for how personal-context AI will likely work everywhere: opt-in, per-action, visible. Third, the English-only launch and tier restrictions mean the real test — whether mainstream users actually talk to their inboxes — won’t begin until availability widens.

For now, the upgrade is live and rolling out this week. If you’re a Google AI subscriber, the fastest way to understand what’s changed is to open Gmail on your phone and just ask it something about your week.