Liquid AI released two open-weight multimodal decision models, d1-3B and experimental d1-omni-600M, for edge deployment
Model release Provisional 86% confidence first seen
Liquid AI released two open-weight multimodal decision models intended to run on edge devices: d1-3B and experimental d1-omni-600M. The models are available on Hugging Face and are described as providing fast, fixed-option probabilistic decisions (rather than generating tokens) using inputs such as text and images, and in one model text+audio as well.