Meta has introduced a handheld AI companion called the Muse Charm, positioning itself in a competitive market alongside devices like the Rabbit R1 and the now-discontinued Humane AI Pin. The device, roughly the size of an Apple Watch, features a teddy bear avatar that interacts through voice commands. Meta has not disclosed pricing or specific launch details beyond a Christmas target, but the product signals a strategic push to put its Muse large language models directly into consumers' hands.

What You Need to Know

This device, roughly the size of an Apple Watch, features a teddy bear avatar and relies on Meta's Muse LLMs. Unlike many competitors, Meta controls the full stack from hardware to AI model. The product is expected to ship by Christmas, but pricing and specific use cases remain unclear.

The Growing Field of AI Companions

Meta's move echoes earlier attempts to create dedicated AI hardware that works alongside or instead of a smartphone. The Rabbit R1 generated buzz but drew criticism for being a pricey gadget that many felt could have been an app. Rabbit itself launched a mobile app this week. OpenAI has tapped Jony Ive to design its first AI handheld, signaling a broader industry race to find a physical form factor for conversational AI. The Muse Charm taps into the nostalgia of the Tamagotchi, a digital pet craze from the 1990s, but swaps virtual care for conversational companionship.

  • Humane AI Pin: A wearable projector that failed to gain traction and was discontinued.
  • Rabbit R1: A bright orange handheld that raised questions about necessity over app-based solutions.
  • Muse Charm: Meta's entry, emphasizing a cute avatar and full-stack control.

Meta's Vertical Integration Advantage

What sets Meta apart from many AI hardware startups is direct control over the hardware, software, and models used. Early AI widgets like the Rabbit R1 relied on off-the-shelf models from the likes of OpenAI. Meta has not confirmed which of its Muse models will power the Charm, but given the device's small size, local processing likely handles voice-to-text and text-to-speech while avatar responses run on cloud-based Muse LLMs. This integrated approach could allow faster iteration and tighter optimization, but it also exposes Meta to the full privacy liability of always-on microphones and cameras. The company has already faced backlash over its Ray-Ban Meta glasses, which earned the moniker pervert glasses after enabling abuse. A new version of those glasses ditches the camera entirely on one model.

Why This Matters

Meta's entry into AI hardware dramatically raises the stakes for startups like Rabbit and OpenAI's upcoming device. With 3 billion users across its platforms, Meta can afford to experiment with pricing and distribution in ways that smaller competitors cannot. The real test will be whether the Muse Charm delivers enough utility to justify carrying a separate device when smartphones already run powerful AI assistants. If Meta succeeds, it could accelerate a shift toward dedicated AI wearables. If it fails, the gadget will join a graveyard of ambitious hardware experiments, but Meta's deep pockets mean the company can afford to iterate on a failed first attempt.