PrismML brings its tiny LLMs to Qualcomm-powered smart glasses
techcrunch ·
PrismML, an AI lab founded by Caltech researchers and advised by UC Berkeley’s Ion Stoica, has developed a version of its tiny language models for smart glasses powered by Qualcomm chips. During Qualcomm’s Snapdragon Summit, the company showcased its 1-bit Bonsai LLM, which is capable of running locally on devices using the Snapdragon AR1 Gen 1 Platform. PrismML specializes in significantly shrinking large models while maintaining high performance on standard benchmarks, with this specific model containing two billion parameters tuned for vision and language tasks. This technology allows users to receive real-time answers about what they are looking at directly on their wearable devices. The startup aims to promote open-weight AI that operates on-device, offering an alternative to proprietary labs that require massive computational resources. By releasing a model optimized for Qualcomm’s hardware, PrismML is advancing its vision of utilizing existing device computing power more effectively. However, as of the announcement, no specific smart glasses products featuring PrismML technology have been officially launched yet.
techcrunch