Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Agent evaluation research benchmarks · SOTA2 Research

Agent evaluation

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Agent Evaluation Dataset (20 agents x 2 requirement types)

0.68Time (min)

10

May 13, 2026

Auto-ClawEval Mini (104 environments)

94.2Safety Score

8

Apr 21, 2026

93.3Safety

8

Apr 21, 2026

FutureX May W2, 56 events

H2O AI Super Agent v1.82

5Rank

6

Jun 23, 2026

OpenThoughts TBLite

72.1Accuracy

5

Jun 24, 2026

VSM full (evaluation)

Claude-Opus 4.6

85.96Average Rating

4

Apr 15, 2026

58.3Accuracy

3

Jun 24, 2026

Terminal-Bench 2.0

40.1Accuracy

3

Jun 24, 2026

FutureX 48 events May W1

1Rank

3

Jun 23, 2026

FutureX Apr. W4 48 events

1Rank

3

Jun 23, 2026

FutureX 56 events Apr. W3

H2O AI Super Agent v1.82

2Rank

3

Jun 23, 2026

86.2Accuracy

2

Jun 24, 2026

Popular datasets

FUTUREX Auto-ClawEval GAIA BFCL Terminal-bench

Follow for updates

Join SOTA2 Contact