← Arena leaderboards

Text · Hard Prompts (English)

Arena · 50 models

RankModelScore
1Gemini 4 Carbon (High)1,545.1093
2Gemini 4 Argon (High)1,539.2419
3Claude Opus 5.5 (High)1,539.1297
4Claude Opus 4.6 (High)1,530.4493
5Claude Opus 4.61,526.8369
6Claude Fable 5.1 (Max)1,523.3689
7Claude Opus 5 (Max)1,523.31
8MiMo-V2.6-Pro1,519.9216
9Claude Opus 5 (High)1,518.4732
10Muse Wonder (Max)1,509.7817
11Claude Fable 5 (High)1,509.5452
12Claude Opus 4.7 (High)1,508.1377
13Qwen3.8-Max1,504.7311
14Gemini 3.8 Flash (High)1,504.1206
15Claude Sonnet 5.5 (xHigh)1,501.447
16Claude Opus 4.71,500.605
17Muse Spark 1.3 (Max)1,499.5747
18GLM-5.3 (Max)1,496.8051
19Kimi K3 (Max)1,496.1153
20MiMo-V2.5-Pro1,493.8724
21Claude Sonnet 4.61,493.285
22Muse Spark 1.2 (xHigh)1,490.5674
23MiMo-V2.6-Flash1,490.1227
24GLM-5.3-Flash1,489.1802
25Gemini 3.7 Flash (High)1,488.6832
26DeepSeek-V4.1-Flash-20260910 (Max)1,488.2532
27Claude Opus 4.8 (High)1,486.9336
28ERNIE 5.11,484.7729
29Gemini 3.6 Flash (High)1,484.7428
30GPT-5.51,483.2678
31GPT-5.5 (High)1,482.5979
32GLM-5.11,482.5085
33Gemini 3.5 Flash (High)1,482.3493
34Qwen3.5-Max-Preview1,481.9233
35Claude Opus 4.5 (20251101)1,481.5947
36GLM-5.2 (Max)1,481.4524
37Muse Spark 1.11,480.9219
38Gemini 3.1 Pro Preview1,480.0108
39Gemini 3.5 Flash (Medium)1,479.1262
40GPT-5.4 (High)1,478.736
41Muse Spark1,477.2369
42GPT-5.6 Sol (xHigh)1,477.1135
43Claude Opus 4.5 (High, 32K, 20251101)1,476.9975
44Claude Opus 4.81,475.4104
45Gemini 3 Pro1,474.7072
46MiMo-V2-Pro1,472.8567
47Qwen3.7-Max-Preview1,472.304
48Claude Sonnet 4.5 (High, 32K, 20250929)1,471.5417
49Kimi K2.61,471.3313
50Claude Sonnet 4.5 (20250929)1,470.0963