Skip to content

Benchmark

Arena ELO Text Factuality - Coding

Coding benchmarks compare models on programming tasks. Higher scores are better.

Models measured
155
Scores recorded
174
Sources
1
IQ weight
1.95

Score distribution

1,491.4
Mean
1,494.7
Median
38
Std dev
1,437
10th percentile
1,534.3
90th percentile
1,344.9
Min
1,344.91,567
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.95 for chat models, measured over 155 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 1.1· LMArena data1,516.9
52Kimi K2.5Instant· LMArena data1,516.9
53Grok 4.5· LMArena data1,516.3
54Claude Fable 5.2Max· LMArena data1,515.9
55Kimi K2.6· LMArena data1,514.8
56Step 5 Preview· LMArena data1,513.8
57GLM 5.2Max· LMArena data1,513.5
58GPT-6.2 SolMax· LMArena data1,512.8
59Gemini 3.5 FlashMedium· LMArena data1,512.2
60Claude Sonnet 5High· LMArena data1,512.1
61GPT-5.6 LunaxHigh· LMArena data1,511.1
62GLM 5.1· LMArena data1,511
63ERNIE 5.1· LMArena data1,510.7
64Claude Opus 4.1Thinking 16K, 20250805· LMArena data1,510.6
65Muse Spark· LMArena data1,506.8
66Grok 4.6High· LMArena data1,506.4
67Gemini 3.1 Pro Preview· LMArena data1,506.3
68Qwen 3.8 27B· LMArena data1,505.3
69MiMo V2 Pro· LMArena data1,504.8
70MiMo V2.5· LMArena data1,504.3
71Claude Opus 4.120250805· LMArena data1,504.1
72DeepSeek V4 Pro· LMArena data1,503.9
73Kimi K2.5Thinking· LMArena data1,503.5
74Inkling FP4· LMArena data1,502.5
75Hy3 (Deep Infra)· LMArena data1,502.1
76Qwen3.7 Plus· LMArena data1,501.9
77MiMo V2 Omni· LMArena data1,500.7
78GPT-6 SolMax· LMArena data1,500.4
79GLM 5· LMArena data1,500.2
80Gemma 4 31B IT· LMArena data1,499.8
81LongCat-Flash-Chat Exp2602· LMArena data1,499.4
82Grok 4.20 Multi-Agent0309, beta· LMArena data1,498.2
83DeepSeek V4 ProHigh, 20260813· LMArena data1,498.1
84Grok 4.20Reasoning, 0309, beta· LMArena data1,497.3
85GPT 5.1High· LMArena data1,496.2
86Qwen3.5 397B A17B· LMArena data1,496.1
87MiniMax M3· LMArena data1,494.7
88Mistral Medium 3.5· LMArena data1,493.9
89DeepSeek V4 Flash· LMArena data1,493.5
90Claude Haiku 4.520251001· LMArena data1,493.4
91Gemini 3.5 Flash Lite· LMArena data1,493.4
92Qwen3.6 Plus· LMArena data1,492.4
93Gemini 3 Flash· LMArena data1,492.1
94DeepSeek V3.2· LMArena data1,491.9
95GPT 5.2High· LMArena data1,491.7
96MiniMax M2.7· LMArena data1,490.8
97Gemini 2.5 Pro· LMArena data1,490.2
98DeepSeek V3.2Thinking· LMArena data1,490
99Mistral Large 4· LMArena data1,489.2
100Mistral Large 3· LMArena data1,488.4

Showing 51–100 of 174 results.

Data credits: LMArena, CC BY 4.0 (licence)