快讯
Liquid AI 发布 Open d1：两款开放权重多模态决策模型，可在边缘设备实时运行
Liquid AI 宣布发布 Open d1，即其 d1 决策模型家族中的两个开放权重多模态模型：d1-3B 支持文本与视觉，d1-omni-600M 支持文本加图像或文本加音频。官方称模型可实现"随时随地实时决策"，部署范围从数据中心（如 NVIDIA DGX）、RTX 工作站到边缘端 Jetson 设备。转发者 Dmitry Sokolowski 称，新模型在小型 NVIDIA Jetson Orin Nano 上回答一个问题约需 50 毫秒，无需云端往返、决策在设备本地完成。
查看原文
RT Dmitry Sokolowski New open models from @liquidai answer a question in about 50 milliseconds on a small @nvidia Jetson Orin Nano. No cloud round trip. The decision happens on the device. Liquid AI: Today we release Open d1: two open-weight multimodal models in our d1 decision model family. > d1-3B: text + vision > d1-omni-600M: text + image or text + audio > Real-time decision making anywhere, from data centers such as @nvidia DGX to RTX workstations to Jetson at the edge.