List your company

Single-step Action Prediction on WebLinx (test-iid)

28.5Cumulative RuntimeCoverage
24.19253.27182.35111.429May 28, 2026

Evaluation results

MethodLinks
CoverageEvaluation approach=Proxy2026.0528.50.6
Qwen3.5-122BEvaluation approach=End-to-end2026.0511742.6
MiniMax-M2.5Evaluation approach=End-to-end2026.05136.249.5