← Arena leaderboards

Text · Factuality · Mathematical

Arena · 50 models

RankModelScore
1Claude Opus 5 (High)1,522.6023
2Gemini 4 Argon (High)1,508.9667
3Claude Fable 5.1 (Max)1,506.2895
4Claude Opus 5 (Max)1,505.5768
5Claude Opus 4.6 (High)1,505.5625
6Claude Opus 4.61,499.3199
7GPT-5.4 (High)1,497.9394
8Claude Fable 5 (High)1,494.008
9GPT-5.5 (High)1,493.0376
10GPT-5.6 Sol (xHigh)1,489.7238
11Gemini 3.8 Flash (High)1,489.633
12Claude Opus 4.7 (High)1,487.2554
13Kimi K3 (Max)1,487.2299
14GPT-6 Astra (Max)1,485.9091
15GPT-5.51,485.7749
16Claude Opus 4.8 (High)1,485.5759
17ERNIE 5.11,483.4457
18Grok 4.51,482.894
19GPT-5.6 Terra (xHigh)1,482.7073
20Gemini 3.7 Flash (High)1,482.6953
21Gemini 3.5 Flash (Medium)1,482.3988
22Claude Opus 4.71,482.2747
23MiMo-V2.5-Pro1,481.0023
24GLM-5.3 (Max)1,480.8975
25Claude Sonnet 5 (High)1,480.5284
26Claude Opus 4.81,480.4778
27Gemini 3.6 Flash (High)1,479.8321
28Muse Spark 1.3 (Max)1,478.4195
29GLM-5.3-Flash1,477.687
30DeepSeek-V4.1-Flash-20260910 (Max)1,475.5645
31Kimi K2.61,475.1142
32GPT-5.41,474.6455
33GLM-5.2 (Max)1,472.4651
34Claude Sonnet 4.61,471.9821
35Inkling1,471.5068
36Gemini 3 Pro1,471.4225
37GPT-5.6 Luna (xHigh)1,470.9056
38Qwen3.8-Max1,470.4032
39Gemini 3.5 Flash (High)1,468.7412
40Claude Opus 4.5 (20251101)1,468.1489
41Gemini 3.1 Pro Preview1,467.9582
42Qwen3.7-Plus1,467.5817
43Hy31,464.0608
44Muse Spark 1.11,463.7761
45Grok 4.6 (High)1,462.5248
46GLM-51,461.1148
47Claude Opus 4.5 (High, 32K, 20251101)1,460.6757
48Kimi K2.5 (Thinking)1,460.4149
49Claude Sonnet 4.5 (High, 32K, 20250929)1,459.4708
50MiMo-V2.6-Flash1,459.3336