← Arena leaderboards

Text · Style Control · Polish

Arena · 50 models

RankModelScore
1Gemini 4 Argon (High)1,525.2078
2Claude Fable 5.1 (Max)1,524.674
3Claude Fable 5 (High)1,512.1358
4Muse Spark 1.11,509.5154
5Gemini 3.5 Flash (High)1,506.0835
6Gemini 3 Pro1,504.6566
7Claude Opus 5 (High)1,504.007
8Claude Opus 4.6 (High)1,502.5293
9Claude Opus 4.71,502.0149
10Gemini 3.7 Flash (High)1,500.6119
11Claude Opus 5 (Max)1,500.126
12Gemini 3.8 Flash (High)1,496.5834
13Gemini 3 Flash1,496.5108
14Gemini 3.5 Flash (Medium)1,495.826
15Gemini 3.1 Pro Preview1,495.3561
16Claude Opus 4.61,493.8627
17GPT-5.6 Sol (xHigh)1,490.5623
18Claude Opus 4.7 (High)1,490.4267
19Muse Spark 1.3 (Max)1,490.1358
20Dola Seed 2.0 Pro1,486.9954
21Grok 4.1 (Thinking)1,486.3225
22Muse Spark1,485.7249
23Claude Opus 4.8 (High)1,485.5546
24ERNIE-5.0-Preview (1203)1,485.1717
25Claude Opus 4.81,483.7978
26GPT-5.4 (High)1,482.7771
27Gemini 3 Flash (Minimal)1,481.0276
28Grok 4.20 Multi-Agent Beta (0309)1,479.888
29GPT-5.5 (High)1,479.7818
30GPT-5.5 Instant1,479.3972
31Kimi K3 (Max)1,478.2114
32Qwen3.8-Max1,478.1348
33Grok 4.7 (xHigh)1,477.8328
34GPT-5.2 Chat (20260210)1,477.2621
35Claude Opus 4.5 (High, 32K, 20251101)1,476.3765
36Grok 4.11,476.0882
37GPT-5.3 Chat1,474.6668
38Grok 4.20 Beta (Reasoning, 0309)1,474.0865
39Gemini 3.6 Flash (High)1,472.9279
40DeepSeek-V4.1-Flash-20260910 (Max)1,472.7999
41GLM-5.3 (Max)1,472.7903
42GPT-5.41,472.408
43GPT-5.51,471.4871
44Claude Sonnet 4.61,471.2542
45Claude Opus 4.5 (20251101)1,470.3145
46MiMo-V2.5-Pro1,467.5437
47Grok 4.20 Beta 11,466.1358
48DeepSeek-V4-Pro1,465.3737
49GPT-5.6 Terra (xHigh)1,465.2289
50ERNIE 5.11,464.3564