NVIDIA 发布 Nemotron 3.5 Lightning：为长时 AI 代理提供快速精准执行
原标题：NVIDIA Nemotron 3.5 Lightning Delivers Fast, Accurate Specialized Task Execution for Long-Running Agents
AI 摘要
NVIDIA 发布了 Nemotron 3.5 Lightning，一个 30B 参数的 MoE 模型，仅有 3B 活跃参数，专为长时间运行的 AI 代理的高频执行任务设计。该模型支持 OpenClaw 和 Hermes Agent 等框架，并配合 NVIDIA NeMo Switchyard 进行模型路由，以及 NeMoClaw 安全与管理栈。Nemotron 3.5 Lightning 在速度和准确性上达到 Pareto 前沿，输出速度比同类模型快 4 倍，在 PinchBench 上以 86% 的准确率比 Qwen3.6 35B 快 30%。模型权重、训练数据和配方以 OpenMDW-1.1 许可发布，支持微调和强化学习。
正文节选
Long-running AI agents spend most of their time on high-volume execution: tool calls, result validation, and subagent delegation. Using a frontier reasoning model for every execution step adds cost and latency. NVIDIA Nemotron 3.5 Lightning is an open 30B mixture-of-experts (MoE) model with 3B active parameters built for that execution layer of always-on agents. It is designed for harnesses like OpenClaw and Hermes Agent—all supported by the NVIDIA NemoClaw open source security and management st