Skip to content

Benchmark

Arena ELO Text Factuality - Spanish

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

Models measured
86
Scores recorded
99
Sources
1
IQ weight
1.96

Score distribution

1,446.3
Mean
1,448.2
Median
24.9
Std dev
1,416.6
10th percentile
1,476.7
90th percentile
1,371.3
Min
1,371.31,500.2
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.96 for chat models, measured over 86 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Qwen 3.8 27B· LMArena data1,446.5
52Kimi K2.5Thinking· LMArena data1,446.1
53Claude Sonnet 4.5High, 32K, 20250929· LMArena data1,442.8
54GPT 5.1High· LMArena data1,442.7
55Claude Opus 4.120250805· LMArena data1,441.5
56Grok 4.6High· LMArena data1,441.5
57Qwen3 235B A22B Instruct 25072507· LMArena data1,441.1
58Gemini 3.5 Flash Lite· LMArena data1,440.4
59Claude Sonnet 5High· LMArena data1,439.8
60Inkling Small· LMArena data1,439.1
61Grok 4.20Reasoning, 0309, beta· LMArena data1,439
62Qwen3.5 397B A17B· LMArena data1,438.8
63Gemini 2.5 Flash· LMArena data1,438.7
64DeepSeek V4 Flash· LMArena data1,437.3
65Gemini 3.1 Flash Lite· LMArena data1,436.6
66Grok 4.1Thinking· LMArena data1,436.4
67ChatGPT-4o Latest20250326· LMArena data1,434.1
68Doubao Seed 2.0 Pro· LMArena data1,431.7
69GPT-6 LunaMax· LMArena data1,430.6
70Grok 4.7xHigh· LMArena data1,430.6
71Mistral Large 3· LMArena data1,429.6
72GPT-6 SolMax· LMArena data1,429.5
73Gemini 3 FlashMinimal· LMArena data1,428.8
74Claude Haiku 4.520251001· LMArena data1,428.5
75DeepSeek V3.2Thinking· LMArena data1,428.1
76GPT 5.1· LMArena data1,428.1
77Qwen3.5 27B· LMArena data1,427.5
78GPT 5.2· LMArena data1,425.8
79Mistral Medium 3.12508· LMArena data1,425.4
80Qwen3.6 Plus· LMArena data1,424.7
81GPT-5.4 miniHigh· LMArena data1,423.8
82MiMo V2.5· LMArena data1,423.7
83MiMo V2 FlashNon-Thinking· LMArena data1,422.4
84MiniMax M3· LMArena data1,421.9
85GPT 5.2High· LMArena data1,421.8
86Gemini 2.5 Flash-Lite PreviewNon-Thinking, 09-2025· LMArena data1,421.8
87DeepSeek V3.2· LMArena data1,419.2
88DeepSeek-V4-Flash PreviewHigh· LMArena data1,417.2
89Grok 4.1· LMArena data1,416.6
90Step 3.5 Flash· LMArena data1,416.1
91Solar Pro 4· LMArena data1,415
92MiniMax M2.7· LMArena data1,414.6
93Grok 4.3· LMArena data1,410.7
94Qwen3.5 Flash· LMArena data1,410
95Kimi K2 Thinking· LMArena data1,401.3
96Grok 4.1 FastReasoning· LMArena data1,398.7
97GPT 5.4 NanoHigh· LMArena data1,398.4
98MiniMax M2.5· LMArena data1,377.8
99Nemotron 3.5 Lightning 30B A3BNVFP4· LMArena data1,371.3

Showing 51–99 of 99 results.

Data credits: LMArena, CC BY 4.0 (licence)