← Arena leaderboards

Text · Factuality · Chinese

Arena · 50 models

RankModelScore
1Claude Opus 5.5 (High)1,557.1478
2Claude Opus 5 (Max)1,555.1585
3Gemini 4 Argon (High)1,551.5966
4Claude Fable 5.1 (Max)1,546.5562
5Claude Opus 4.6 (High)1,544.9955
6Claude Opus 5 (High)1,543.5782
7Claude Opus 4.61,542.2615
8Claude Fable 5 (High)1,540.5769
9GPT-5.6 Sol (xHigh)1,539.4928
10GPT-6 Astra (Max)1,537.9683
11GPT-5.51,537.5151
12Gemini 3.7 Flash (High)1,535.8101
13GPT-5.5 (High)1,534.7049
14GPT-5.4 (High)1,532.9832
15Gemini 3 Pro1,532.7981
16Gemini 3.8 Flash (High)1,532.5916
17Qwen3.8-Max1,528.3793
18Claude Opus 4.7 (High)1,527.5671
19Claude Opus 4.71,527.3659
20GPT-5.6 Terra (xHigh)1,526.6194
21Gemini 3.5 Flash (High)1,525.1529
22GPT-5.41,525.1399
23Gemini 3.1 Pro Preview1,523.9383
24Gemini 3.5 Flash (Medium)1,523.7668
25GLM-51,523.5036
26Kimi K3 (Max)1,523.3243
27GLM-5.3 (Max)1,523.3015
28Gemini 3.6 Flash (High)1,521.0301
29GLM-5.3-Flash1,520.9677
30Claude Opus 4.81,517.9885
31Kimi K2.61,517.1637
32Gemini 3 Flash1,516.9793
33GLM-5.2 (Max)1,516.7657
34Muse Spark 1.3 (Max)1,515.9357
35GLM-4.61,515.6297
36GPT-6.1 Sol (Max)1,514.67
37GLM-5.11,514.4192
38DeepSeek-V4.1-Flash-20260910 (Max)1,514.2551
39Claude Opus 4.8 (High)1,513.9423
40MiMo-V2.5-Pro1,511.7637
41Qwen3.7-Plus1,511.4774
42GPT-5.1 (High)1,509.7144
43MiMo-V2.6-Flash1,506.9655
44Grok 4.51,506.6469
45ERNIE 5.11,505.9183
46Muse Spark 1.2 (xHigh)1,504.6635
47Muse Spark 1.11,502.54
48Qwen3.8-27B1,501.746
49DeepSeek-V4-Pro Preview (High)1,501.2721
50Qwen3.5-397B-A17B1,500.9931