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

Role-Play Evaluation on Minimax Role-Play Bench

84.65Average Score

MiniMax-M2-her

43.809254.412165.01575.6179Jan 29, 2026
Updated 3mo ago

Evaluation Results

MethodLinks
84.6580.5579.9797.51-
2026.01
80.6376.6272.2197.05-
76.6267.2382.189.9-
75.662.7283.8793.08-
69.3555.7275.6690.28-
68.2352.3682.1186.08-
2026.01
66.3964.9646.2389.4-
2026.01
65.7359.1357.7486.9-
64.2251.1166.4588.21-
2026.01
61.2550.6659.5384.15-
2026.01
60.7247.2756.6591.71-
60.2745.8166.6482.83-
2026.01
58.4447.2952.7886.4-
2026.01
57.6343.3250.1193.78-
2026.01
50.7640.3832.8289.48-
2026.01
48.4729.8747.5186.64-
2026.01
45.3834.3230.3282.58-