Meta’s Muse AI Now Calls and Video Chats With You as an Animated Avatar

Follow Teck Hustlers on Google Add Teck Hustlers as a preferred source to see more of our stories in Google Top Stories. Add as […]

Get a summary in:
Glowing digital avatar face, AI assistant concept
Follow Teck Hustlers on Google
Add Teck Hustlers as a preferred source to see more of our stories in Google Top Stories.
Add as preferred source on Google

Meta is rolling out “Muse Voice” to everyone in the US, letting people call and video chat with its consumer AI agent Muse as a talking, animated avatar instead of just texting it.

Meta announced the expansion Thursday, pushing Muse Voice out to the full US user base after first showing it off at Connect in September 2026. The feature runs on two new models, Muse Realtime Avatar and Muse Realtime Voice, and defaults to a character nicknamed “Jolly” that you can reshape into a person, an animal, or something else entirely. While it talks, the avatar throws in small expressions and gestures that track the mood of the conversation. It works on mobile, web, and Mac right now, with glasses and a dedicated wearable gadget still to come.

Smartphone video call app interface

What Meta actually shipped

Muse has existed as a text-based assistant for a while. Muse Voice turns it into something you talk to out loud and watch react in real time, rather than something you type at and read back from. Meta frames this as a shift from a chat window to a conversation partner with a face.

The two models doing the work are Muse Realtime Avatar, which handles the visual side (the expressions, the gestures, the lip movement), and Muse Realtime Voice, which handles the actual back-and-forth speech. Together they’re meant to make the avatar feel present during a call rather than just narrating responses at you.

Jolly is the default face you get, but Meta built in customization so the avatar can look like a person, an animal, or whatever character someone picks. That flexibility matters to Meta’s bigger bet here, which is that the look of the thing shapes how attached people get to it.

Why Meta is doing this

Meta laid out its thinking in a research blog post titled “bringing-your-muse-to-life.” The argument is straightforward: giving Muse a face and a voice, anthropomorphizing it, could make people feel more connected to it, and that connection gets stronger the more live conversations someone has with it.

That’s a deliberate design choice, not a side effect. A text box is easy to close. A video call with something that looks back at you and reacts to what you say is a different kind of interaction, and Meta is betting that difference translates into people coming back more often and staying longer when they do.

It’s also part of a broader pattern this week of Meta pushing its consumer AI and social products harder. The company’s Instagram arm just rolled out a desktop launch for Edits, its video editing app, another sign Meta is trying to get more of its everyday tools in front of more people on more devices at once.

Alexandr Wang’s dog, and the joke that came with it

Meta’s chief AI officer Alexandr Wang posted his own demo of Muse Voice on X, showing off a Muse he’d customized to look like a fluffy white dog. In the clip, Wang joked that his habit of posting memes on X has landed him in trouble with his boss more than once. His customized Muse, apparently picking up on that running joke, suggested he start adding disclaimers to memes that could be misread as Meta endorsing something risky.

It’s a small moment, but it’s doing real promotional work. Having the company’s own AI chief show his avatar giving him lighthearted reputational advice is a soft way of demonstrating that these avatars can hold a natural, specific conversation rather than just reciting canned responses. It’s also a reminder that a chattier, friendlier-looking AI agent is still making judgment calls about tone and risk, the same kind of judgment call that matters a lot more once these agents start handling tasks with actual consequences. That’s the same territory CrowdStrike was circling with its recent acquisitions of Fiddler and Dreadnode, buys aimed squarely at securing AI agents as they take on more autonomous, human-facing roles.

Where this is headed next

Muse Voice works now on mobile, web, and Mac. Meta says real-time conversations are coming next to its AI Glasses, folding the avatar into a device people already wear on their face for long stretches of the day.

More notable is “Muse Charm,” a small wearable Meta describes as Tamagotchi-like, with a small screen that displays the avatar so people can talk to it wherever they are, not just when they’re sitting in front of a phone or laptop. Meta plans to launch Muse Charm before the end of the year, which puts it on a tight runway from here.

Putting the avatar on a dedicated wearable is the clearest signal of what Meta wants Muse to become. A chat app is a tool you open when you need something. A device you wear that displays a face and talks back is designed to feel like company, not utility. That’s a very different product category, closer to a pet or a companion than a search bar with a personality.

The bigger AI assistant race

Meta isn’t alone in pushing voice into its AI products. Other AI competitors offer voice chat features of their own, and the push toward spoken, more natural interaction with AI agents has been building across the industry for a while now. What sets Muse Voice apart is the visual layer on top of the voice, an avatar with a face and body language, not just a voice in your ear.

That combination raises the stakes on getting this right. A voice assistant that gives a wrong answer is annoying. An animated character that someone has spent months growing attached to, that then gives bad advice, mishandles a sensitive conversation, or gets manipulated into saying something it shouldn’t, is a different kind of problem. The more human these agents look and sound, the more scrutiny they deserve, not less.

Meta hasn’t said anything about usage numbers, pricing, or whether Muse Voice is coming to markets outside the US. For now this is a US-only rollout, built around getting as many existing Meta users as possible to try talking to Jolly, or whatever version of it they design for themselves, before the Charm wearable shows up later this year.

The bottom line

Meta is making a calculated bet that a face and a voice create loyalty a chat window can’t, and the Charm wearable is the real tell: this isn’t about better customer service, it’s about building something people miss when they put it down. That’s a powerful retention mechanic, and also exactly the kind of product that deserves more scrutiny than Meta’s research blog post gives it. Making an AI agent lovable and making it trustworthy are two different engineering problems, and right now Meta has only talked publicly about solving the first one.

Follow Teck Hustlers on Google
Add Teck Hustlers as a preferred source to see more of our stories in Google Top Stories.
Add as preferred source on Google
About

The Teck Hustlers Editorial Team covers technology, business, finance and gaming news for readers who want the story behind the headline. Every article is researched, written and edited by our in-house team, checked against company announcements, filings and reputable reporting, and updated when new facts emerge. We focus on what each development means for founders, professionals, investors and everyday users. Have a tip or a correction? Reach us through our contact page.

Up next: Serve Robotics Wants to Rent Out Its Robot Brains and Become the AWS of Robotics