Skip to content

Benchmark

Arena ELO - Hard Prompts English

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

Models measured
364
Scores recorded
421
Sources
1
IQ weight
1.31

Score distribution

1,319.1
Mean
1,374
Median
157.7
Std dev
1,085.3
10th percentile
1,479.2
90th percentile
734.7
Min
734.71,547.3
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.31 for chat models, measured over 364 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
101DeepSeek V3.2· LMArena data1,445.6
102GLM 4.6· LMArena data1,444.9
103Mistral Medium 3.5· LMArena data1,444.9
104Grok 4.6High· LMArena data1,444.7
105LongCat Flash Chat· LMArena data1,444.2
106DeepSeek V3.2 ExpThinking· LMArena data1,443.8
107Grok 4.20 Beta 1· LMArena data1,443.5
108Mistral Large 4· LMArena data1,442.4
109Grok 3 Preview02-24· LMArena data1,441.6
110GPT 5.1· LMArena data1,440.2
111MiniMax M2.7· LMArena data1,440.2
112Gemini 3.5 Flash Lite· LMArena data1,440.1
113Grok 4.1· LMArena data1,440.1
114Qwen3 VL 235B· LMArena data1,439.2
115DeepSeek-V4-Flash PreviewHigh· LMArena data1,438.6
116Mistral Large 3· LMArena data1,437.7
117DeepSeek V3.2 Exp· LMArena data1,437.5
118Hunyuan Vision 1.5 Thinking· LMArena data1,436.2
119Qwen3 235B A22B Instruct 25072507· LMArena data1,436.2
120ERNIE-5.0-Preview1203· LMArena data1,436.2
121MiMo V2 FlashNon-Thinking· LMArena data1,436.2
122Grok 4.1Thinking· LMArena data1,435.5
123Claude Haiku 4.520251001· LMArena data1,435
124Mistral Medium 3.12508· LMArena data1,434.6
125Kimi K2 Thinking· LMArena data1,433.1
126Gemini 3 FlashMinimal· LMArena data1,432.4
127nova-experimental-chat26-01-10· LMArena data1,432.3
128Qwen 3 235B A22B2507, thinking· LMArena data1,432.3
129DeepSeek V3.1 TerminusThinking· LMArena data1,431.6
130Qwen3.5 122B-A10B· LMArena data1,431.6
131Qwen3 Next 80B A3B Instruct· LMArena data1,431.3
132GPT 5.2High· LMArena data1,430.6
133NVIDIA Nemotron 3.5 Super VL 120BNVFP4· LMArena data1,429.9
134GLM 4.5· LMArena data1,429.8
135Qwen3 Max2025-09-23· LMArena data1,429.1
136Inkling Small· LMArena data1,429
137ChatGPT-4o Latest20250326· LMArena data1,428.4
138Qwen3.5 27B· LMArena data1,427.9
139Solar Pro 4· LMArena data1,427.5
140DeepSeek V3.1Thinking· LMArena data1,427.1
141GPT-6.2 AstraMax· LMArena data1,425.7
142nova-experimental-chat11-10· LMArena data1,424.5
143Inkling 2 Small· LMArena data1,424.1
144GPT-5.5 Instant· LMArena data1,423
145GPT 5.2· LMArena data1,422.7
146DeepSeek R10528· LMArena data1,421.7
147Grok 4.7xHigh· LMArena data1,420.5
148GPT-5.4 miniHigh· LMArena data1,420.4
149Claude Haiku 5.5High· LMArena data1,420.3
150Gemini 2.5 Flash· LMArena data1,420.2

Showing 101–150 of 421 results.

Data credits: LMArena, CC BY 4.0 (licence)