Liquid AI releases open d1 decision models d1-3B and d1-omni-600M
Liquid AI released two open decision models: d1-3B and experimental d1-omni-600M. d1-3B supports text and images; d1-omni-600M supports text and images or text and audio for edge inference.
What changed
Released two open d1 decision models: d1-3B (text and images) and experimental d1-omni-600M (text+image or text+audio).
Why it matters
Developers can run multimodal decision models on edge GPUs with low latency; d1-3B runs under 50 ms on measured Jetson devices.
Who should care
DevelopersResearchersEnterprises
Worth trying
Yes — Open-weight on Hugging Face; requires transformers>=5.14.