← Arena leaderboards

Text · Style Control · Hard Prompts

Arena · 50 models

RankModelScore
1Gemini 4 Carbon (High)1,548.1746
2Gemini 4 Argon (High)1,544.3494
3Claude Opus 5.5 (High)1,534.7645
4Claude Opus 4.6 (High)1,530.6
5Claude Fable 5 (High)1,529.336
6Claude Opus 4.61,525.8045
7Claude Opus 4.7 (High)1,525.2642
8Muse Wonder (Max)1,524.1982
9Claude Fable 5.1 (Max)1,523.6935
10Claude Opus 4.71,518.8432
11Gemini 3.8 Flash (High)1,517.0327
12Kimi K3 (Max)1,516.8174
13Claude Opus 5 (High)1,516.3225
14Muse Spark 1.3 (Max)1,516.057
15Claude Opus 5 (Max)1,514.5498
16Muse Spark 1.2 (xHigh)1,513.4069
17Claude Opus 4.8 (High)1,513.3362
18GPT-5.6 Sol (xHigh)1,510.9813
19Muse Spark 1.11,510.5905
20Gemini 3.7 Flash (High)1,509.1568
21Gemini 3.1 Pro Preview1,507.6212
22GPT-6.1 Sol (Max)1,506.3691
23Muse Spark1,505.8939
24Claude Opus 4.81,504.9045
25Claude Sonnet 5.5 (xHigh)1,504.8781
26GLM-5.3 (Max)1,504.7221
27MiMo-V2.6-Pro1,504.3394
28Qwen3.8-Max1,503.5839
29Claude Sonnet 4.61,503.1184
30Gemini 3 Pro1,502.2757
31GLM-5.3-Flash1,501.782
32GPT-5.5 (High)1,501.296
33Gemini 3.6 Flash (High)1,501.2369
34Claude Opus 4.5 (High, 32K, 20251101)1,500.7325
35GPT-6 Astra (Max)1,499.8516
36GPT-5.51,498.9225
37DeepSeek-V4.1-Flash-20260910 (Max)1,498.7269
38Claude Opus 4.5 (20251101)1,498.0744
39GPT-5.2 Chat (20260210)1,496.6329
40GPT-5.4 (High)1,496.5242
41GLM-5.2 (Max)1,496.1631
42Gemini 3.5 Flash (Medium)1,495.2964
43Gemini 3.5 Flash (High)1,495.131
44Qwen3.7-Max-Preview1,494.703
45MiMo-V2.5-Pro1,494.1855
46Gemini 3 Flash1,492.5908
47Qwen3.5-Max-Preview1,491.1521
48GPT-5.5 Instant1,491.1043
49Claude Sonnet 5 (High)1,490.5464
50GPT-5.6 Terra (xHigh)1,489.5275