New · November 5, 2026
Muse Voice is live in the US — talk to your Muse, watch it react
On November 5, Meta rolled out Muse Voice to everyone in the US — real-time voice conversations with an expressive avatar that reacts as you talk. Tap the phone icon in the top-right corner of the Muse app and you're talking, not typing. It was previewed at Connect in September; now it's the shipped version of the biggest interaction change since Muse launched.
What shipped
🎙️ Real-time voice
Two new models power it: Muse Realtime Voice (immediate spoken responses) and Muse Realtime Avatar (the face). Meta's pitch is thinking out loud: "Sometimes it's easier to think out loud. With voice, you can talk through an idea with your Muse and let it take shape without worrying about the exact prompt." You can interrupt mid-sentence, the way you would with a person, and Muse keeps working on the task in the background while you talk — on a walk, making dinner, wherever.
🙂 An avatar that reacts
As your Muse speaks, the avatar uses "subtle expressions and gestures" that match the tone of the conversation — it reacts while you talk, so you can tell it's following along. The default avatar is nicknamed Jolly, and you can customize it into a person, an animal, or another character. You choose how present it is: voice only, the avatar alongside your conversation, or full screen when you want to talk face to face. One honest limit: Muse doesn't see you — it hears you and reacts based on the conversation (TechCrunch).
🎭 Hundreds of voices
Meta offers hundreds of voices to choose from — "the energy of a New Yorker or the flair of a pirate" — and you can change it whenever you want. The launch demo came from Meta chief AI officer Alexandr Wang, who showed his own Muse as a fluffy white dog, joking about his meme-posting habit getting him in trouble with his boss; his Muse suggested adding disclaimers to memes that could be misread as Meta endorsing something risqué.
Where it works — and where it doesn't yet
- US only, for now. Rolling out in the Muse app in the US on mobile, web and Mac. Meta has given no date for Canada, Japan, Singapore or Mexico — the four other markets where Muse is live.
- Glasses and Charm are next, not now. Meta says voice conversations are "coming soon" to AI glasses and Muse Charm, no date. Zuckerberg said on the Oct 21 earnings call that Muse reaches Meta's glasses starting in November, and Charm is expected before year-end — but whether voice lands on glasses on the same schedule as Muse itself is not established. See Muse on glasses.
- It was a gradual rollout first. Days before the announcement, users reported live voice quietly reaching the app over the weekend (Nov 1), subscribers and early-access users first, with no Meta announcement at the time. The Nov 5 announcement made it official for everyone in the US.
- Same day, second launch. Meta also launched Muse in Mexico on Nov 5 — its fifth market and first in Latin America. See Muse in Mexico.
The honest read
- Voice changes what Muse is for. Everything before this was texting an agent that runs errands. Voice turns it into a conversation you can have while doing something else — and the avatar is Meta betting that a face makes you trust the agent more. TechCrunch's framing: anthropomorphizing Muse could make consumers feel more connected to their AI agent, and that connection could grow stronger as they begin to engage in live conversations.
- The rehearsal use case is the real one. Meta's own examples — prepare for a job interview, run through a presentation, talk through a pitch — are where voice beats typing: hearing a response and answering in the moment makes it feel more like the real thing. That's also the use case where an interruption-friendly, tone-reacting avatar actually matters.
- Watch the glasses version. Voice on glasses is the Connect promise that matters most — hands-free, wake-word Muse — and it's still "coming soon". If the November glasses launch ships without voice, the first version is a text agent on your face. The glasses page tracks that: Muse on glasses.
- No price change announced. Voice is part of the Muse app rollout in the US; Meta hasn't said it's a paid-tier feature, and no subscription price was mentioned in the announcement. The $20/$100 paid tiers and free tier stand as before.
5 prompts to try voice with
Open a voice chat (phone icon, top right) and talk — these are built for speaking, not typing. All US-only for now.
The half-baked trip
I'm thinking out loud about a trip — the city I've always wanted to see is Kyoto, I want to wake up somewhere quiet with a garden view, and there's one meal I'd plan the whole trip around: proper ramen. Talk me through it as I think, ask me questions when I'm vague, and start building the plan while we talk.
Interview rehearsal, live
Pretend you're the hiring manager for a product manager role. Ask me one question at a time, let me answer out loud, then tell me what was strong and what was vague before the next question. Keep the pace like a real interview — don't wait for me to type.
The presentation run-through
I'm about to pitch [TOPIC] to my team in 10 minutes. Let me run the opening 60 seconds out loud, then stop me, tell me where I lost the thread, and give me a tighter version. Then run it again with me.
Voice-only vs. avatar, compared
Do this task twice and tell me the difference: first, help me plan my week voice-only, no avatar. Then do the same plan with the avatar on screen. Which version caught more of my hesitations, and did seeing a face change how much detail I gave you?
Pick a voice that fits the job
I need two voices: one for a focused work session (calm, direct, no jokes) and one for brainstorming (energetic, playful, pushes back). Suggest which of your voice styles fits each, switch between them as I ask, and tell me if the voice changes how I talk to you.
Good to know
- Timeline: Previewed at Connect Sept 23, 2026 (voice mode + Realtime Avatar) · users reported a quiet rollout starting Nov 1 · officially rolled out to all US users Nov 5 · glasses and Charm still "coming soon".
- The dance button (Nov 3): two days before Voice, Wang announced you can make your Muse avatar dance on command and share the clip to Instagram, Facebook or WhatsApp (or save it to your phone). Same avatar, different trick — the avatar is becoming a shareable character, not just a chat face.
- What Meta calls it: "Muse Voice" for the feature, "Muse Realtime Avatar" and "Muse Realtime Voice" for the two models behind it. The avatar's nickname is Jolly.
- The context: voice was one of the four Connect-era features still marked "not shipped" as of Oct 31 (along with the dedicated agent email, glasses, and the Confidential VM). Email and Confidential VM are still waiting. See Muse on glasses and Muse in Mexico — same-day launch.
- New here? Set up Muse in 15 minutes → · Get 1B free tokens →
← Muse on glasses · Muse in Mexico → · All news
Not affiliated with Meta. Sources: Meta newsroom, "Give Your Muse a Voice" (Nov 5, 2026); TechCrunch, "Meta says US users can now video chat with their Muse AI avatar" (Nov 5, 2026); Alexandr Wang's Nov 5 X post and demo; agentstacklist's Nov 1 voice-mode rollout report and Nov 3 dance-feature note. Glasses and Muse Charm voice dates are unannounced.