PrismML, an AI lab founded by Caltech researchers and advised by UC Berkeley’s Ion Stoica, has adapted its compact language models to run on smart glasses powered by Qualcomm’s Snapdragon chips. At Qualcomm’s Snapdragon Summit, the company demonstrated its 1-bit Bonsai large language model (LLM) operating locally on AI smart glasses built on the Snapdragon AR1 Gen 1 Platform.

PrismML’s technology is notable for significantly reducing the size of large language models—by a factor of four—while maintaining comparable performance on standard benchmarks. The version designed for smart glasses features a 2-billion-parameter model fine-tuned for both vision and language tasks, allowing users to query their surroundings in real time.

The startup aims to promote open-weight AI models that run efficiently on edge devices, leveraging existing hardware capabilities rather than relying on cloud-based AI services. This approach also addresses privacy concerns associated with proprietary AI platforms that require extensive computational resources.

While PrismML’s collaboration with Qualcomm marks progress toward integrating advanced AI into wearable devices, no commercial smart glasses featuring PrismML’s models have been announced to date.