← Arena leaderboards

Text · Style Control · Japanese

Arena · 50 models

RankModelScore
1Claude Fable 5.1 (Max)1,553.3959
2Claude Opus 5.5 (High)1,534.5796
3Gemini 4 Argon (High)1,533.0562
4Claude Fable 5 (High)1,517.1922
5Gemini 3.7 Flash (High)1,514.6497
6Kimi K3 (Max)1,508.5507
7Claude Opus 5 (Max)1,507.4606
8Claude Opus 5 (High)1,505.8868
9GPT-5.5 (High)1,504.6335
10Gemini 3.8 Flash (High)1,504.2674
11Gemini 3 Pro1,503.3595
12Claude Opus 4.6 (High)1,497.9137
13Gemini 3.1 Pro Preview1,496.2744
14GPT-5.6 Sol (xHigh)1,495.2344
15Gemini 3 Flash1,488.2675
16Claude Opus 4.7 (High)1,488.0165
17Qwen3.5-Max-Preview1,483.7481
18Gemini 3.5 Flash (High)1,482.3006
19Muse Spark 1.3 (Max)1,482.047
20Claude Opus 4.61,479.9451
21GPT-5.51,478.9655
22GPT-5.4 (High)1,477.9964
23Gemini 3.6 Flash (High)1,477.1321
24Grok 4.20 Beta 11,476.3989
25Qwen3.8-Max1,475.1999
26Muse Spark 1.2 (xHigh)1,473.2481
27Claude Opus 4.71,472.6265
28GLM-5.3 (Max)1,467.3621
29Muse Spark 1.11,465.2993
30Claude Sonnet 5.5 (xHigh)1,464.7255
31GPT-5.6 Terra (xHigh)1,462.9508
32GPT-5.5 Instant1,462.7388
33GPT-5.1 (High)1,460.6617
34Claude Opus 4.8 (High)1,459.5796
35Claude Opus 4.81,459.5171
36Gemini 3.5 Flash (Medium)1,459.1124
37GLM-5.3-Flash1,456.9996
38DeepSeek-V4-Pro1,455.0555
39GLM-5.2 (Max)1,454.5413
40DeepSeek-V4.1-Flash-20260910 (Max)1,452.4904
41GPT-5.2 Chat (20260210)1,448.0475
42GPT-6.1 Sol (Max)1,447.5346
43Claude Opus 4.5 (High, 32K, 20251101)1,446.8844
44Gemini 3.5 Flash-Lite1,446.1608
45GPT-5.41,446.0425
46Gemini 2.5 Pro1,443.8845
47Grok 4.51,441.6641
48Claude Sonnet 5 (High)1,441.64
49Kimi K2.61,440.1198
50GLM-5.11,440.1166