OR-Clarify Benchmark for LLM-Driven Optimization Modeling
September 3, 2026
OR-Clarify is a new benchmark evaluating how well LLMs identify missing objectives and constraints in natural-language operations research requests. It measures an agent's ability to perform bounded, interactive clarification before formulating mathematical programs.
HOW THIS AFFECTS YOU
●
builderYou can use this benchmark to evaluate the reliability of LLM agents in automated optimization workflows.
●
researcherThis introduces a formal way to measure pre-formulation clarification capabilities in LLMs.