According to recent industry reports, social media and AI giant Meta is aggressively expanding its footprint in the smart hardware sector. Meta plans to sell 10 million wearable devices in the second half of this year through new product launches and market expansion, aiming to reach 6.8 million monthly active users (MAUs) for its wearables by year-end. This ambitious sales target signals Meta's full commitment to edge-AI hardware following the initial success of its Ray-Ban Meta smart glasses. Furthermore, Meta plans to officially initiate testing for a novel "AI Pendant" next year, further diversifying the physical forms of its AI Agent ecosystem.
Alongside the hardware acceleration, the capital market for large language models (LLMs) is also witnessing major developments. MiniMax, a leading Chinese general LLM unicorn, officially signed a tutorial agreement with CITIC Securities on May 29, formally launching its A-share IPO process. As one of China's top AI startups, MiniMax has established a strong presence in social and productivity AI applications. This pioneering IPO move marks a critical transition for Chinese AI unicorns from pure technological R&D to accelerated capitalization and commercialization.
[AgentUpdate Depth Analysis] Meta's 10 million sales target and AI pendant roadmap highlight the inevitable evolution of AI Agents from pure software interfaces to embodied hardware entities. Compared to the struggles of pioneers like Humane AI Pin and Rabbit R1, Meta holds a significant advantage by leveraging mature form factors (like Ray-Ban glasses) and its vast social ecosystem. The AI pendant, as a low-barrier, ambient device, will serve as a critical gateway for multimodal AI Agents to perceive the physical world in real-time via cameras and microphones. Moving forward, the competition in the AI Agent space will shift from cloud-based LLM parameter wars to a battle for edge-hardware entry points and real-time contextual awareness. Meta's aggressive strategy is rapidly ushering in the era of "Hardware as an Agent."
