Meta Muse AI Agent Gets Voice Chat Mode With Hundreds of Voices

On Friday, November 6, 2026, Meta announced real-time voice chat for its Meta Muse AI assistant. The update turns the consumer-facing agent into something users can call, with reportedly hundreds of distinct voices to choose from.

What the Voice Chat Mode Does

The change is easy to spot: a phone icon now appears in the top-right corner of the Muse conversation screen. Tap it to start a live call. The agent listens, responds aloud, and keeps working on the request as you talk; you can interrupt it mid-sentence, just as you might on a phone call, and it adjusts its response.

Voice selection is the headline feature. The agent reportedly offers hundreds of distinct voice options, and Meta has kept the picker in the existing conversation interface rather than moving it to a separate settings menu. Meta hasn’t published an exact breakdown of voice categories; coverage of the announcement cites the size of the voice library. The feature is reportedly available only in the Muse app on mobile, web, and Mac, and only in the United States.

Voice and Avatar Technology Behind the Update

Meta says two parts of its own technology make the exchange feel conversational rather than scripted: Muse Realtime Voice handles speech, while Muse Realtime Avatar provides the on-screen presence. Reports say the avatar’s lip movements and facial expressions sync with its words, and its face reacts as it listens to the user.

That distinction matters. A voice mode with a static waveform is already a solved problem in this category; the competition now centers on latency, interruption handling, and the social cue of a face responding to you. Meta presents Muse as an agent you delegate work to, rather than a search box you query. A call interface reinforces that division of labour: you talk, and it executes.

Context: Two Months After Launch

Muse has been publicly available for about two months, and voice mode is at least the third visible change in that time. Since release, Meta has shipped two additional Muse Wonder models alongside earlier platform upgrades, giving the product one of the faster early release cadences among current AI assistants.

The timing puts Meta in a familiar race. Voice-first interaction has become table stakes across the category, and rivals have taught users to expect low-latency calls that handle interruptions. Meta is betting that voice and an expressive avatar, paired with an agent that completes tasks, feel like a different kind of product from a chatbot that talks back. It’s still early enough, at two months in, to reposition that bet.

What Happens Next

Geographic expansion is the obvious next step. A US-only launch on mobile, web, and Mac leaves major international markets waiting, and Meta’s history with feature rollouts points to staged, country-by-country availability rather than a global switch. Regional pricing and packaging for any premium tier remain unconfirmed.

Platform coverage is another area to watch. Android availability, third-party app integrations, and Application Programming Interface (API) access for developers building on Muse could each broaden where people use voice mode. Meta hasn’t shared a timeline for any of these steps, so the near-term signal to watch is whether the phone icon appears outside the US before the year closes. Bottom line: Meta has made its agent something you call, not just something you type to. With the launch limited to the US, the next announcement is about where it arrives, not what it does.

Read Next

• Instagram Expands Edits Video Editor to macOS Desktops
• Samsung Android XR Smart Glasses Pre-Orders Start November 12
• Claude Personal Finance Tool Can Analyze Your Spending

FAQs