New Delhi: Meta has introduced a new feature that allows users in the United States to have real-time voice and video conversations with its animated artificial intelligence assistant, Muse. The update moves beyond traditional text-based AI interactions by giving the assistant a digital face that responds with facial expressions, gestures and synchronised lip movements.

The new feature, called Muse Voice, is available across mobile devices, the web and Mac computers in the US. Instead of typing questions and waiting for text responses, users can now interact with an animated character through live voice and video calls.

Meta’s Chief AI Officer Alexandr Wang announced the rollout on Thursday through a company blog post and a post on X. The launch represents another step in Meta’s efforts to make AI assistants more interactive and personalised.

Muse AI gets an animated face

Unlike conventional voice assistants that communicate through audio alone, Meta’s Muse comes with an animated digital avatar.

The default character is nicknamed Jolly. It uses subtle facial expressions and gestures that change according to the conversation, making interactions appear more natural.

When a user speaks to Muse, the character listens and responds through voice while displaying corresponding expressions. The avatar is designed to create the experience of talking to a digital character rather than interacting with a conventional chatbot.

However, Meta has clarified that the AI cannot actually see the user. Its reactions are based on what the user says during the conversation rather than visual information captured from the user’s surroundings.

The feature is intended to make AI conversations feel more personal and engaging, particularly for people who prefer speaking instead of typing prompts.

Users can customise their AI avatar

One of the notable features of Muse Voice is the ability to customise the appearance of the digital assistant.

Users are not restricted to the default Jolly character. Meta allows them to change the avatar’s appearance to resemble a different person, an animal or a custom character.

This gives users the option to create an AI companion with a visual identity that suits their preferences.

During a demonstration shared on X, Alexandr Wang showed a customised version of Muse designed as a fluffy white dog. He interacted with the character and joked about whether his memes could get him into trouble at work.

The demonstration highlighted how Meta is positioning Muse as a personalised digital companion rather than simply another AI tool for answering questions.

Meta believes that giving an AI assistant a face and personality can make conversations feel more natural and encourage users to engage with it more frequently.

Two new AI models power real-time conversations

The technology behind Muse Voice relies on two AI models developed by Meta: Muse Realtime Avatar and Muse Realtime Voice.

These models work together to create a synchronised experience during live conversations.

Muse Realtime Voice handles the spoken interaction, while Muse Realtime Avatar helps generate the character’s visual expressions and movements.

One of the main technical challenges in animated AI conversations is keeping the character’s lip movements aligned with its spoken words. Delays or mismatched expressions can make the interaction appear unnatural.

Meta says the two models are designed to keep the avatar’s voice, facial expressions and lip movements synchronised during real-time calls.

The company first previewed the video conversation feature at its Connect developer conference in September. The latest rollout brings that demonstration closer to everyday consumer use.

Meta adds watermarking and age restrictions

Along with the new conversational features, Meta has introduced measures intended to address safety and transparency.

All video conversations generated through the system carry an invisible watermark called Meta Video Seal. The watermark is designed to remain embedded in generated video content and help identify material created using Meta’s AI technology.

Meta has also restricted access to Muse to users aged 18 and above.

The age restriction is particularly relevant as AI companions become more interactive and begin to resemble human conversations. The company has not positioned Muse Voice as a service for children or younger teenagers.

These measures form part of Meta’s approach to managing the risks associated with AI-generated visual and voice content.

Muse Charm wearable device planned for launch

Meta’s plans for Muse extend beyond smartphones and computers. The company is preparing a small wearable device called Muse Charm, which it intends to launch before the end of 2026.

The device is described as a palm-sized gadget resembling a digital pet, similar to the Tamagotchi toys that became popular in the late 1990s.

Muse Charm is expected to feature a small screen displaying the user’s customised AI avatar. This would allow people to interact with Muse and ask it to complete tasks without having to take out their smartphones.

The concept could make AI assistants more accessible throughout the day by placing them in a dedicated wearable device.

Meta has also indicated that real-time video conversation support will eventually be introduced to its smart glasses. This could allow users to interact with their AI assistant through wearable technology rather than relying entirely on mobile screens.

However, the wearable device and smart-glasses integration are future plans, not features confirmed as widely available with the current Muse Voice rollout.

Meta moves towards more human-like AI interaction

The introduction of Muse Voice reflects a broader shift in the development of consumer AI assistants. Technology companies are increasingly exploring ways to make interactions more natural through voice, visual characters and real-time responses.

Traditional chatbots mainly rely on written prompts and text-based answers. Animated AI assistants attempt to create a more conversational experience by combining spoken communication with facial expressions and visual personalities.

For Meta, Muse could become part of a wider ecosystem connecting AI services across mobile devices, computers, wearable gadgets and smart glasses.

The US rollout gives users an opportunity to experience this new form of interaction, although availability remains geographically limited. Meta has not announced an India launch date in the report.

As the company expands its AI offerings, the success of Muse will depend on how effectively it combines useful assistance, natural conversations, privacy protections and safeguards for users.