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

Multi-turn Conversation on MT-Bench (Win Rate)

77.1Win Rate

RPO

49.74856.84963.9571.051Jun 16, 2025
Updated 1d ago

Evaluation Results

MethodLinks
2025.06
77.1
2025.06
76.8
2025.06
76.4
2025.06
75.9
2025.06
75.1
2025.06
74.6
2025.06
68.2
2025.06
67.4
2025.06
64.3
2025.06
63.8
2025.06
61.5
2025.06
61
2025.06
60.9
2025.06
60.3
2025.06
60.2
2025.06
59.1
2025.06
58.8
2025.06
57.9
2025.06
57.4
2025.06
56.9
2025.06
56.8
2025.06
56.2
2025.06
56.1
2025.06
56
2025.06
55.9
2025.06
55.8
2025.06
55.4
2025.06
55.2
2025.06
53.7
2025.06
53.5
2025.06
52.9
2025.06
52.9
2025.06
52.8
2025.06
52.1
2025.06
51.5
2025.06
50.8