Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Long-context Evaluation on LongBench

52.6Average Score

FullKV

20.890429.122737.35545.5873Feb 9, 2026Feb 14, 2026Feb 19, 2026Feb 24, 2026Mar 1, 2026Mar 6, 2026Mar 12, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
52.6
2026.03
50.67
2026.03
49.28
2026.03
48.92
2026.03
48.39
2026.03
47.47
2026.03
46.95
2026.03
46.45
2026.03
46.4
2026.03
46.19
2026.03
46.17
2026.03
46.12
2026.03
45.7
2026.03
45.61
2026.03
44.82
2026.03
44.4
2026.03
44.31
2026.03
44.22
2026.03
44.02
2026.03
43.82
2026.03
43.82
2026.03
43.78
2026.03
43.68
2026.03
43.23
2026.03
43.02
2026.03
42.79
2026.03
42.47
2026.03
42.27
2026.03
42.22
2026.03
42.03
2026.03
41.88
2026.03
41.81
2026.03
41.33
2026.03
40.96
2026.03
40.78
2026.03
40.09
2026.03
39.98
2026.03
38.9
2026.03
38.45
2026.02
37.97
2026.02
37.63
2026.02
36.5
2026.02
36.13
2026.03
31.96
2026.03
31.63
2026.03
31.39
2026.03
31.32
2026.03
31.14
2026.03
31.01
2026.03
30.62
2026.03
26.76
2026.03
26.06
2026.03
26.04
2026.03
25.88
2026.03
25.36
2026.03
25.33
2026.03
22.11