Meta opens video chats with its AI agent Muse to everyone in the US as Muse Voice goes live
Meta has switched on Muse Voice for all users in the United States, letting them call and video chat with its personal AI agent through a real-time voice conversation and an expressive animated avatar.
Meta has opened Muse Voice to everyone in the United States, the company announced on Thursday, 5 November, turning its consumer AI agent Muse from a text-based assistant into a live voice and video companion that users can talk to in real time.
Muse Voice lets users call and video chat with Muse through the Muse app on mobile, on the web and on Mac. The conversation is spoken rather than typed: users tap the phone icon in the top-right corner of the app and start talking, and they can interrupt Muse mid-answer, much as they would cut in during a phone call with a person.
The experience is built on two new AI models, Muse Realtime Voice and Muse Realtime Avatar. The voice model produces immediate spoken responses, while the avatar model generates an animated character that reacts with facial expressions and gestures matching the tone of the conversation as it listens and speaks.
That character has a default form — a cute animated avatar nicknamed Jolly — but users can change its appearance to resemble a person, an animal or another character. Meta says hundreds of voices are available to choose from, including styles inspired by a New Yorker or a pirate, and the selection can be changed at any time.
The avatar does not actually see the user. It hears them and reacts on the basis of the conversation, so the exchange is a spoken dialogue with a visual response rather than a video call with another person. Meta says the generated avatar video carries a durable, invisible Meta Video Seal watermark.
The rollout was announced in a blog post and on X by Meta's chief AI officer, Alexandr Wang, who shared a demonstration of his own Muse in action — a fluffy white dog that joked with him and offered tongue-in-cheek advice about meme-posting etiquette.
Meta first previewed the video-chat feature at its Connect developer conference in September 2026. Thursday's announcement is the step from preview to general availability in the United States, rather than the first introduction of the avatar technology itself.
Muse itself was introduced on 8 September 2026 as Meta's personal AI agent, built around messaging and delegated action — sending emails, shopping, planning trips and handling tasks on the user's behalf through the Muse app or WhatsApp. Voice adds a spoken, visual interface to an agent that began as a text conversation, and Meta says voice conversations can keep running in the background while tasks continue.
The launch comes just before Meta's plan to release a Tamagotchi-like wearable device for Muse before the end of the year. The device, called Muse Charm, would show the avatar on a small screen. Meta says real-time conversations will later come to its AI Glasses and to Muse Charm, but neither is part of the initial rollout, which covers mobile, web and Mac only.
The service is for users aged 18 and older, and Meta has not named any countries beyond the United States for the app. Whether the combination of live voice, an expressive avatar and background task-running changes how people use a personal AI agent — and whether Meta can extend it to glasses and a dedicated wearable — is the open question the rollout will now answer.
More in Technology & AI
Photo: Anthropic chief executive Dario Amodei in 2023 (file image)
Technology & AIAnthropic lets Claude read your bank account with new Personal Finance tool — but it cannot move a cent
Anthropic has launched Personal Finance in Claude, a read-only beta that links US bank and card accounts through Plaid so users can ask where their money goes — while the chatbot stays unable to move, manage or invest a cent.
Photo: Tux, the Linux penguin mascot (file image)
Technology & AIOpenAI's AI Agents Build 948 Working Root Exploits for the Linux Kernel in a Single $1.2 Million Run
OpenAI says an autonomous pipeline built on an upcoming cyber model reproduced 1,724 Linux kernel vulnerabilities and turned 948 of them into working root exploits, each with a candidate patch, for $1.21 million in API credits.
Photo: The TikTok app on a smartphone screen (file image)
Technology & AIPTA dismisses fake TikTok ban notification as ‘false and misleading’, says no order issued
The Pakistan Telecommunication Authority said on Thursday that no notification banning TikTok had been issued, after a fabricated document claiming a nationwide suspension spread across social media.