Instruction Following Evaluation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
86.69IFEval Score
32
May 13, 2026
76Score
24
May 29, 2026
22Delta Win Rate (ΔWR)
24
Feb 26, 2026
73.2Score
23
May 29, 2026
86.1Score
23
May 29, 2026
83.7Accuracy
18
May 12, 2026
51.9GPT-4o Score
17
Apr 7, 2026
51.6GPT-4o Score
17
Apr 7, 2026
49.9GPT-4o Score
17
Apr 7, 2026
51.8GPT-4o Score
17
Apr 7, 2026
48.14Win Rate
16
Mar 6, 2026
78.06HSR
14
Jun 23, 2026
38ArenaHardv1 Score
14
Mar 4, 2026
54.17LC% over π0
10
Mar 4, 2026
71.8Quality
9
Jun 4, 2026
62.17Length-Controlled Win Rate
8
Jun 25, 2026
87Score
6
Apr 30, 2026
56.73Score
5
Feb 26, 2026
74.29Score
5
Feb 26, 2026
69.48Score
5
Feb 26, 2026
72.01Score
5
Feb 26, 2026
71.29Score
5
Feb 26, 2026
72.06Score
5
Feb 26, 2026
58.5A Win Rate
5
Feb 26, 2026
62A Win Rate
5
Feb 26, 2026