Meta Adds Real‑Time Voice Mode to Muse, Turning Its AI Agent into a Talking Companion

Meta has switched Muse, its AI assistant, into full conversational gear with a new real‑time voice mode on iOS. Announced by CEO Mark Zuckerberg on November 5, 2026, the feature is now rolling out to users in the U.S., turning Muse from a text‑first chatbot into something closer to a talking companion.

With voice mode enabled, you simply tap a microphone icon and speak naturally to Muse. The assistant transcribes and processes your speech in real time, replying out loud while an animated avatar appears at the top of the screen. The avatar reacts to the flow of conversation, adding facial expressions and motion that make the interaction feel more like a video call than a static chat window. Because no typing is required, the feature is particularly suited to situations like commuting, cooking, or walking between meetings.

Early hands‑on reports note that responses are generally quick, with occasional pauses when Muse looks up information on the web. Reviewers say the avatar can feel slightly uncanny at times, but agree that it meaningfully boosts the sense of presence compared with text‑only interfaces. Meta has also confirmed that voice mode is part of a broader hardware push: the company plans to bring the same capability to its AI glasses and an upcoming “Muse Charm” wearable, extending spoken interaction beyond the smartphone. For now, though, the rollout is limited to the U.S. Muse app on iOS, with timing for Android and web support still unclear.