PrismML ports tiny open-weight LLMs to Qualcomm’s Snapdragon AR1 Gen 1 for smart glasses

PrismML, a startup founded by Caltech researchers and advised by UC Berkeley’s Ion Stoica, has adapted its tiny language models to run on Qualcomm’s Snapdragon AR1 Gen 1 Platform for smart glasses. At Qualcomm’s Snapdragon Summit, the chipmaker demonstrated PrismML’s 1-bit Bonsai LLM running locally on AI smart glasses hardware. The model is a 2-billion-parameter variant tuned for vision and language, enabling wearers to ask questions about what they see in real time without sending data to the cloud.

The startup positions its approach as a way to shrink larger models substantially—by 4× in this case—while retaining almost all of their performance on standard benchmarks. PrismML’s broader goal is open-weight AI that runs efficiently on existing device silicon, offering an alternative to proprietary cloud-based AI services that demand ever-increasing compute resources.

Qualcomm’s showcase highlights the integration of PrismML’s model with its Snapdragon AR1 Gen 1 chip, which targets augmented-reality wearables. Despite the demonstration, no specific smart glasses products running PrismML have been announced yet, and the companies did not provide a timeline for commercial availability.