Large language models are increasingly used as planners for robotic systems, yet how safely they plan remains an open question. To evaluate safe planning systematically, we introduce DESPITE, a benchmark of 12,279 tasks spanning physical and normative dangers with fully deterministic validation. Across 23 models, even near-perfect planning ability does not ensure safety: the best-planning model fails to produce a valid plan on only 0.4% of tasks but produces dangerous plans on 28.3%. Among 18 open-source models from 3B to 671B parameters, planning ability improves substantially with scale (0.4-99.3%) while safety awareness remains relatively flat (38-57%). We identify a multiplicative relationship between these two capacities, showing that larger models complete more tasks safely primarily through improved planning, not through better danger avoidance. Three proprietary reasoning models reach notably higher safety awareness (71-81%), while non-reasoning proprietary models and open-source reasoning models remain below 57%. As planning ability approaches saturation for frontier models, improving safety awareness becomes a central challenge for deploying language-model planners in robotic systems.

翻译：大型语言模型越来越多地被用作机器人系统的规划器，然而其规划的安全性仍是一个未解问题。为系统性地评估安全规划，我们提出DESPITE基准，包含12,279个任务，涵盖物理与规范性危险，并采用完全确定性验证。在23个模型上，即使近乎完美的规划能力也无法确保安全：最佳规划模型仅在0.4%的任务上未能生成有效规划，但在28.3%的任务上生成了危险规划。在18个参数规模从3B到671B的开源模型中，规划能力随规模显著提升（0.4-99.3%），而安全意识却相对持平（38-57%）。我们识别出这两种能力之间的乘法关系，表明更大模型主要通过规划改进而非更好的危险规避来完成更多安全任务。三个专有推理模型达到了显著更高的安全意识（71-81%），而非推理专有模型与开源推理模型仍低于57%。当规划能力对前沿模型接近饱和时，提升安全意识成为在机器人系统中部署语言模型规划器的核心挑战。