← Arena leaderboards

Text · Style Control · Hard Prompts (English)

Arena · 50 models

RankModelScore
1Gemini 4 Carbon (High)1,543.6893
2Gemini 4 Argon (High)1,542.8419
3Claude Opus 5.5 (High)1,537.9693
4Claude Opus 4.6 (High)1,535.0089
5Claude Fable 5 (High)1,531.2315
6Claude Opus 4.61,529.4997
7Claude Opus 4.7 (High)1,528.4372
8Muse Wonder (Max)1,527.3113
9Claude Fable 5.1 (Max)1,521.1385
10Claude Opus 4.71,520.2354
11Kimi K3 (Max)1,517.1258
12Claude Opus 4.8 (High)1,516.9894
13Muse Spark 1.3 (Max)1,515.1101
14Gemini 3.8 Flash (High)1,513.7371
15MiMo-V2.6-Pro1,513.4961
16Claude Opus 5 (High)1,513.4911
17GPT-6.1 Sol (Max)1,512.9217
18Claude Sonnet 5.5 (xHigh)1,512.9194
19Claude Opus 5 (Max)1,512.7746
20GLM-5.3 (Max)1,511.2597
21Muse Spark 1.2 (xHigh)1,510.5942
22Qwen3.8-Max1,510.0992
23GPT-5.6 Sol (xHigh)1,509.3371
24Muse Spark1,508.3771
25GPT-6 Astra (Max)1,507.0809
26Gemini 3.1 Pro Preview1,506.6718
27Claude Sonnet 4.61,506.1679
28Claude Opus 4.81,506.1547
29Muse Spark 1.11,505.3065
30DeepSeek-V4.1-Flash-20260910 (Max)1,504.2133
31Claude Opus 4.5 (High, 32K, 20251101)1,503.2128
32MiMo-V2.5-Pro1,500.2263
33Gemini 3.6 Flash (High)1,500.0117
34GPT-5.51,499.8518
35GPT-5.2 Chat (20260210)1,499.7416
36Gemini 3 Pro1,499.6714
37GLM-5.3-Flash1,499.517
38Gemini 3.7 Flash (High)1,499.3826
39GLM-5.11,499.3553
40Claude Opus 4.5 (20251101)1,499.1244
41GPT-5.5 (High)1,498.8537
42GLM-5.2 (Max)1,498.6565
43Claude Sonnet 4.5 (High, 32K, 20250929)1,494.6794
44Claude Sonnet 5 (High)1,493.9648
45Qwen3.5-Max-Preview1,493.792
46GPT-5.4 (High)1,493.059
47Grok 4.51,492.7707
48ERNIE 5.11,491.7127
49Gemini 3 Flash1,491.0142
50Gemini 3.5 Flash (Medium)1,490.9814