← Arena leaderboards

Text · Style Control · Mathematical

Arena · 50 models

RankModelScore
1GPT-6.1 Sol (Max)1,538.8387
2Gemini 4 Argon (High)1,538.3828
3Claude Fable 5.1 (Max)1,536.0332
4Claude Opus 5 (High)1,532.2408
5Claude Opus 4.6 (High)1,524.1356
6Claude Opus 5 (Max)1,522.8129
7Claude Fable 5 (High)1,521.2011
8Gemini 4 Carbon (High)1,518.9973
9Claude Opus 5.5 (High)1,518.9156
10Claude Sonnet 5.5 (xHigh)1,515.5995
11Claude Opus 4.7 (High)1,515.4498
12Gemini 3.8 Flash (High)1,512.5559
13GPT-5.6 Sol (xHigh)1,511.4164
14Claude Opus 4.61,511.3634
15Muse Wonder (Max)1,508.9729
16Kimi K3 (Max)1,508.0469
17GPT-6 Astra (Max)1,507.4142
18Claude Opus 4.8 (High)1,506.4867
19GPT-5.5 (High)1,506.3867
20MiMo-V2.6-Pro1,503.7968
21GLM-5.3 (Max)1,500.2787
22GPT-5.51,500.1893
23Muse Spark 1.3 (Max)1,499.1508
24GPT-5.4 (High)1,498.9614
25Muse Spark 1.11,498.4403
26Claude Opus 4.71,497.6944
27Muse Spark 1.2 (xHigh)1,497.555
28Gemini 3.7 Flash (High)1,497.4889
29DeepSeek-V4.1-Flash-20260910 (Max)1,495.3799
30GLM-5.3-Flash1,495.0141
31Qwen3.8-Max1,494.9232
32Grok 4.51,494.3529
33Hy31,494.1525
34GLM-5.2 (Max)1,494.1306
35GPT-5.6 Terra (xHigh)1,490.5739
36MiMo-V2.5-Pro1,490.5599
37Claude Sonnet 5 (High)1,489.3567
38ERNIE 5.11,489.068
39Gemini 3.5 Flash (Medium)1,488.87
40Gemini 3.6 Flash (High)1,488.8473
41Qwen3.6-Max-Preview1,488.7779
42Claude Opus 4.81,488.7734
43Beam 1.0 501B A23B1,487.9609
44Qwen3.5-Max-Preview1,486.419
45Gemini 3.1 Pro Preview1,485.3762
46Grok 4.6 (High)1,484.9592
47Gemini 3.5 Flash (High)1,484.8326
48GPT-5.6 Luna (xHigh)1,484.8294
49Kimi K2.61,484.1363
50Claude Sonnet 4.61,483.0643