Open d1: Edge decision models for text, vision, and audio

d1-3B scores 48.57 on the Decision Index v0.2.1, matching Decider 35B-A3B despite being 12x smaller. The open-weight decision model, trained from LFM2.5-VL-3B, answers a question in 8 ms on an NVIDIA GeForce RTX 4090 and under 50 ms on Jetson edge devices. Alongside it, d1-omni-600M, built from LFM2.5-Encoder-350M, handles text with image or audio inputs. Both are available on Hugging Face with day-one llama.cpp support.