Skip to content

Benchmark

Arena ELO - Chinese

People compare model answers and choose which they prefer. Higher scores are better.

Models measured
340
Scores recorded
396
Sources
1
IQ weight
1.23

Score distribution

1,334.6
Mean
1,384.4
Median
172.2
Std dev
1,067.9
10th percentile
1,512.4
90th percentile
719
Min
7191,628.9
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.23 for chat models, measured over 340 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Muse Spark· LMArena data1,506.9
52MiMo V2.5 Pro· LMArena data1,506.6
53Claude Opus 4.8High· LMArena data1,506
54GPT-6.1 SolMax· LMArena data1,504
55DeepSeek V4.1 Flash 0910Max· LMArena data1,503.4
56GLM 4.6· LMArena data1,500
57ERNIE-5.0-Preview1203· LMArena data1,499.8
58Nemotron 3 UltraNVFP4· LMArena data1,499.6
59Kimi K3.1 ProMax· LMArena data1,498.8
60Qwen3.5 397B A17B· LMArena data1,497.5
61Gemma 4 26B A4B· LMArena data1,495.6
62Step 5 Preview· LMArena data1,495.1
63Grok 4.5· LMArena data1,495.1
64GPT 5.1High· LMArena data1,494.7
65GLM 4.7· LMArena data1,494.3
66GPT-6 AstraMax· LMArena data1,493.9
67Kimi K2.5Thinking· LMArena data1,493.8
68Inkling FP4· LMArena data1,493.5
69ERNIE 5.00110· LMArena data1,493.3
70Hy3 (Deep Infra)· LMArena data1,492.1
71Qwen 3.8 27B· LMArena data1,491.1
72Claude Sonnet 4.6· LMArena data1,489.7
73Doubao Seed 2.0 Pro· LMArena data1,489.5
74DeepSeek-V4-Pro PreviewHigh· LMArena data1,486.5
75Qwen3.6 Max Preview· LMArena data1,485.6
76Grok 4.20Reasoning, 0309, beta· LMArena data1,482.1
77Qwen 3 235B A22B2507, thinking· LMArena data1,481.3
78DeepSeek V4 Pro· LMArena data1,480.6
79Qwen3.5 27B· LMArena data1,478.7
80Qwen3 Max Preview· LMArena data1,478.4
81DeepSeek V4 ProHigh, 20260813· LMArena data1,477.8
82Qwen3.6 Plus· LMArena data1,476.5
83Gemini 3 FlashMinimal· LMArena data1,476
84Grok 4.20 Multi-Agent0309, beta· LMArena data1,476
85Claude Sonnet 5High· LMArena data1,475.8
86Grok 4.20 Beta 1· LMArena data1,475.7
87Gemma 4 31B IT· LMArena data1,475.7
88Grok 4.6High· LMArena data1,474.2
89GPT-5.6 LunaxHigh· LMArena data1,472.1
90MiMo V2.5· LMArena data1,471.1
91Claude Opus 4.520251101· LMArena data1,470.1
92Gemini 3.5 Flash Lite· LMArena data1,469.6
93MiMo V2 Omni· LMArena data1,469
94DeepSeek-V4-Flash PreviewHigh· LMArena data1,468.5
95DeepSeek V3.1Thinking· LMArena data1,468
96Claude Opus 4.5High, 32K, 20251101· LMArena data1,467.3
97GPT 5.1· LMArena data1,467.1
98GPT-6.1 TerraMax· LMArena data1,465.8
99Grok 4.1· LMArena data1,465.8
100Mistral Large 4· LMArena data1,465.7

Showing 51–100 of 396 results.

Data credits: LMArena, CC BY 4.0 (licence)