Benchmark RadarAI BENCHMARK PROFILE
OmniRouting
OmniRouting evaluates LLMs on PCB routing reasoning under real-world constraints with 1,681 industrial designs across four tasks: geometric routing, design-rule-aware routing, electrical functionality, and tool-augmented agentic routing.
- Released
- 2026-08-05
- Readiness
- Paper only
- Primary field
- Industrial & Engineering
Why it matters
Current benchmarks do not cover routing under strict geometric, topological, and electrical constraints. OmniRouting fills this gap, exposing limitations in path planning and rule adherence for LMMs in EDA.
Motivation
Recent large language models (LLMs) have demonstrated remarkable progress in constraint-aware navigation, maze reasoning, and graph reasoning.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.