← Arena leaderboards

Text · Factuality · Korean

Arena · 50 models

RankModelScore
1Claude Opus 5 (Max)1,521.2316
2Claude Opus 5 (High)1,493.649
3Claude Opus 4.61,464.0768
4Claude Opus 4.71,463.4868
5GPT-5.5 (High)1,460.8785
6Gemini 3.1 Pro Preview1,458.0711
7GPT-5.6 Sol (xHigh)1,456.2369
8Gemini 3 Pro1,453.5671
9GPT-5.51,451.5791
10Claude Opus 4.6 (High)1,449.8154
11Gemini 3.5 Flash (Medium)1,443.6548
12GPT-5.41,439.1699
13DeepSeek-V4-Pro Preview (High)1,436.6757
14Muse Spark 1.11,436.0425
15Claude Opus 4.81,436.0175
16GPT-5.6 Terra (xHigh)1,433.7476
17GLM-5.2 (Max)1,433.1898
18Claude Opus 4.7 (High)1,431.5552
19Claude Sonnet 5 (High)1,430.4132
20Kimi K3 (Max)1,426.3835
21GLM-5.3-Flash1,425.1306
22Gemini 3 Flash (Minimal)1,422.1621
23Kimi K2.5 (Thinking)1,419.6267
24DeepSeek-V4-Pro1,418.2867
25Gemini 2.5 Pro1,418.1142
26Grok 4.20 Beta (Reasoning, 0309)1,417.6744
27MiMo-V2.5-Pro1,415.6298
28Claude Opus 4.5 (20251101)1,414.8083
29GLM-5.11,413.2363
30Claude Sonnet 4.5 (20250929)1,412.8038
31Grok 4.6 (High)1,409.1647
32Qwen3-235B-A22B-Instruct (2507)1,405.9038
33Dola Seed 2.0 Pro1,405.0791
34Grok 4.11,404.0071
35Claude Sonnet 4.61,402.84
36Gemini 3.1 Flash-Lite Preview1,401.7167
37Inkling1,401.4448
38GPT-5.21,400.2082
39Qwen3.5-397B-A17B1,398.7395
40GPT-6 Luna (Max)1,394.8526
41Gemini 2.5 Flash1,391.9645
42GPT-5.4 mini (High)1,388.5152
43MiMo-V2.51,385.8657
44Mistral Large 31,383.5979
45MiniMax-M31,382.4423
46Grok 4.1 (Thinking)1,382.1857
47Qwen3.6-Plus1,381.7076
48Claude Sonnet 4.5 (High, 32K, 20250929)1,377.3586
49DeepSeek-V4-Flash1,377.0497
50GPT-5.2 (High)1,375.7284