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

Multi-turn Dialogue on MT-Bench (Speedup, Acceptance Length, Drafting Cost Share)

3.22Speedup

EAGLE-3

0.52641.22571.9252.6243May 8, 2026May 11, 2026May 14, 2026May 18, 2026May 21, 2026May 24, 2026May 28, 2026
Updated 5d ago

Evaluation Results

MethodLinks
2026.05
3.226.1631
2026.05
3.185.96-
2026.05
3.185.73-
2026.05
3.126.57-
2026.05
3.14.4616
2026.05
3.096.6-
2026.05
3.075.33-
2026.05
3.065.99-
2026.05
3.025.64-
2026.05
3.015.19-
2026.05
3.016.04-
2026.05
2.976.02-
2026.05
2.965.51-
2026.05
2.926.04-
2026.05
2.915.05-
2026.05
2.875.13-
2026.05
2.866.13-
2026.05
2.835.33-
2026.05
2.815.62-
2026.05
2.85.84-
2026.05
2.85.73-
2026.05
2.775.03-
2026.05
2.764.66-
2026.05
2.734.77-
2026.05
2.75.69-
2026.05
2.664.3126
2026.05
2.55.66-
2026.05
2.44.2329
2026.05
2.363.218
2026.05
2.333.3214
2026.05
2.213.2111
2026.05
2.184.0924
2026.05
2.063.1914
2026.05
1.993.0911
2026.05
1.933.9329
2026.05
1.882.596
2026.05
1.853.8225
2026.05
1.53.4614
2026.05
1.484.2530
2026.05
1.382.4538
2026.05
1.273.5527
2026.05
1.062.679
2026.05
0.92.137
2026.05
0.632.0140