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

Average Performance across Context Lengths on Long-context benchmarks

45.9Performance (8k Context)

L2A

39.76441.35742.9544.543Mar 18, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
45.940.736.534.229.6
2026.03
45.741.437.63529.8
2026.03
45.741.636.230.722.8
2026.03
42.339.434.223.212.8
2026.03
42.239.334.126.922.7
2026.03
4239.432.426.822.4
2026.03
41.729.723.219.717
2026.03
4036.433.53126.1