List your company

ResearchFormerly WizWand

Single-step action prediction

Benchmarks

DatasetSOTA methodMetricTrendResultsLast updated

Cloud Console Benchmark 400-sample

Gemini 3 Pro Preview

95.25Accuracy

8

Jun 9, 2026

WebLinx (test-iid)

28.5Cumulative Runtime

3

May 29, 2026