Metas agent Muse adds voice support and expressive avatars
Meta's AI agent Muse now supports voice chat and expressive avatars in the US across mobile, web, and Mac, so users can talk out loud, pick from hundreds of voices, and watch the Jolly mascot match facial expressions to speech—metas agent muse adds a more natural way to rehearse interviews, pitches, and everyday questions.
Key Takeaways
- Meta is rolling out Muse voice support and expressive avatars in the US on mobile, web, and Mac.
- Users can speak instructions or questions and choose from hundreds of customizable voices.
- Expressive avatars make the Jolly mascot mirror facial expressions while Muse talks.
- Meta positions the feature for interview practice, presentation run-throughs, and live pitch drills.
- Muse passed five million downloads after its September launch, and US users can try the update today.
The update lands as Meta leans hard into a character-first AI pitch. Muse's Jolly mascot already shows up across the company's branding, and voice mode pushes that personality from a silent companion into something you can interrupt, answer, and rehearse with out loud. For a deeper look at how today's tech echoes earlier digital habits, browse our Nostalgia: Then & Now coverage.
What did Meta announce for Muse today?
According to Mashable's report, Meta said Muse now offers voice support so people can talk to their agent when giving instructions or asking questions. The company also introduced expressive avatars alongside that audio rollout.
In Meta's own wording, the company is rolling out voice and expressive avatars in the Muse app across the mobile app, web, and Mac app starting today in the United States. The pitch is a "new, natural way" to interact with Muse rather than typing every prompt.
Voice customization is a centerpiece. Meta says users can pick from hundreds of voices and switch whenever their mood changes, teasing options with "New Yorker energy" or "Pirate flair." That framing matters because it treats the agent less like a single default assistant and more like a cast of speaking personas.
Mark Zuckerberg also teased the update on X, underscoring how central Muse has become to Meta's consumer AI story. Mashable notes that Muse quickly reached more than five million app downloads after launching in September, which helps explain why Meta is iterating in public so fast.
How does Muse voice mode change talking to AI?
Then, most people met AI through text boxes: type a prompt, wait, read a reply. Now, Muse's voice mode asks users to speak, listen, and respond in the moment. Meta's blog examples lean into that shift. The company says some things are better rehearsed out loud, including preparing for a job interview, running through a presentation, or talking through a pitch.
Hearing a response and answering live, Meta argues, makes practice feel closer to the real thing. That is a meaningful product claim even if the underlying model is familiar: the interface becomes a conversation drill, not only a search-and-summarize tool.
The nostalgia angle writes itself. Early smartphone voice assistants often felt stiff, transactional, and easy to abandon after a few failed requests. Muse arrives in an era when AI companies treat talk-and-listen loops as the default path to stickiness. Meta is betting that hundreds of voices plus a mascot people already recognize will make spoken interaction feel playful enough to keep using.
There is a cautionary parallel in the wider industry. Mashable points out that voice mode can go sideways quickly, citing viral videos from online creator Husk that showed ChatGPT's voice mode acting "very, very dumb." Muse still has to prove its spoken replies stay useful under the same public stress tests.
Why do expressive avatars matter for AI agents?
Voice alone changes how you prompt. Expressive avatars change how the agent feels present. Mashable reports that Muse's mascots will match facial expressions to what they are saying, so the Jolly character is not a static sticker while audio plays.
That detail fits a broader industry lesson Mashable highlights: a cute little character is a powerful way to anthropomorphize AI, making the product seem cutesy rather than intimidating. Meta already has momentum there. The Jolly mascot has become a common sight in the company's branding, and expressive face-matching doubles down on that emotional shortcut.
Then, AI often looked like a blank chat window. Now, Muse is packaging intelligence as a speaking figure with mood, voice variety, and visible expression. Whether that makes the product more trustworthy or merely more entertaining is an open question, but it is clearly Meta's chosen path for mainstreaming the agent.
Mashable also notes that Meta's latest AI models have caught up to the frontier amid the company's massive AI spending. Voice and avatars are the consumer packaging around that model race: features people can feel immediately, even if the hard tech gains happen under the hood.
Who can use Muse's new voice features right now?
US users can experiment with Muse's voices and expressive avatars starting today, Mashable reports. Availability is described across Muse's mobile app, web experience, and Mac app, which means the update is not locked to a single device class.
If you already downloaded Muse during its post-launch surge past five million installs, the practical next step is simple: open the app in the US and look for the new spoken interaction and avatar expression tools. If you are new to Muse, the timing is notable—Meta is shipping personality features while the product is still early and culturally loud.
The competitive backdrop remains noisy. Other AI products have already trained audiences to expect voice chats, and public failures travel as fast as demos. Meta's advantage, at least for now, is a recognizable mascot, rapid download growth, and a clear rehearsal-focused use case list.
For readers tracking how yesterday's chatbots became today's talking companions, Muse's update is a clean then-and-now marker: text first, then voice, then a face that moves with the words. The feature set is live in the US today, and the next test is whether those hundreds of voices stay charming after the novelty wears off.