Inside the Growing Market for AI Wearables: From Smart Pendants to Ambient Companions

Form factors are also shifting toward jewelry. Sandbar recently introduced its $299 Stream ring, scheduled to launch this summer in gold and silver finishes. Unlike always-listening pendants, the ring leaves its microphone disabled until the user touches a flat panel on top, allowing users to record voice notes, interact with an AI assistant through a phone app, and control audio playback.

Companions, personal reflection, and pocket assistants

Beyond meeting transcription, several devices explore ambient emotional support and conversational presence. Friend, a $129 white pendant that connects to a smartphone via Bluetooth, monitors the wearer's tone and mood to send proactive text reactions. Following public criticism and vandalized subway ads in New York City referencing surveillance concerns, the company introduced "Friend 2.0," which incorporates an integrated speaker to vocalize its responses directly.

Other companies are pursuing self-improvement rather than social interaction. Nirva, founded by former Meta Reality Labs engineers, plans to release a $249 pendant and bracelet hardware package by late 2025. Instead of transcribing business meetings, Nirva monitors spoken interactions to compile a reflective journal focused on communication habits and emotional trends. Meta itself previewed the Muse Charm, a pocket- or keychain-sized device featuring a compact screen displaying "Jolly," an animated avatar that represents its Muse voice assistant.

Budget and multimodal form factors round out the category. Omi offers an $89 wearable that can be worn around the neck or adhered to the side of the head with medical tape. The device routes spoken audio through ChatGPT to answer questions, build task lists, and track conversation context. For standalone use, Rabbit continues to iterate on its $199 R1 handheld, pairing a rotating camera and touchscreen with its "Creations" tool, which lets users generate custom software and mini-games.

What it means for developers

For software engineers, the surge in AI wearables highlights how specialized hardware increasingly serves as an edge audio collection layer for cloud-based inference pipelines. Most of these devices possess minimal on-board compute; their core utility depends entirely on low-latency streaming to transcription engines, retrieval-augmented generation (RAG) databases, and large language models.

As hardware teams implement real-time contextual awareness—such as Omi feeding speech streams directly into ChatGPT—engineering teams must balance accuracy, context window limitations, and ongoing operational costs. When designing backend microservices for ambient or voice-enabled products, developers can try top AI models cheaply through one API at https://apixoai.online, allowing them to test and compare performance across models like GPT, Claude, and Gemini under continuous streaming workloads.

Additionally, the contrasting approaches between always-listening hardware and push-to-activate systems—such as Sandbar's ring panel or Bee's manual mute control—demonstrate that developers must prioritize clear user consent states and local buffering logic. As consumer scrutiny regarding continuous audio capture grows, backend architectures will need strict access controls and transparent data retention policies to maintain developer and user trust.

Source: The hottest AI wearables and gadgets you should know about — TechCrunch AI. Written by the Apixo team from that report.