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
151Step 3.5 Flash· LMArena data1,420.2
152DeepSeek V3.1· LMArena data1,419.9
153GPT 5High· LMArena data1,419.3
154MiniMax-M2.1 Preview· LMArena data1,418.3
155Grok 4 FastNon-Reasoning· LMArena data1,418
156GPT-6 SolMax· LMArena data1,417.4
157MiMo V2 FlashThinking· LMArena data1,417.3
158Qwen3 VL 235Bthinking· LMArena data1,416.7
159Grok 4.1 FastReasoning· LMArena data1,414.7
160Grok 4 FastReasoning· LMArena data1,413.2
161GPT-6.2 SolMax· LMArena data1,412.6
162nova-experimental-chat10-20· LMArena data1,412.3
163ERNIE-5.0-Preview1022· LMArena data1,411.9
164Qwen3.5 35B A3B· LMArena data1,411
165Gemini 2.5 Flash Preview09-2025· LMArena data1,410.5
166OpenAI o32025-04-16· LMArena data1,410.4
167Grok 40709· LMArena data1,410.3
168Qwen3.5 Flash· LMArena data1,408.8
169GPT-6 LunaMax· LMArena data1,407.4
170GPT-4.5 Preview2025-02-27· LMArena data1,406.6
171Qwen3 30B A3B Instruct2507· LMArena data1,406.2
172DeepSeek V3.1 Terminus· LMArena data1,406
173Claude Opus 4Thinking 16K, 20250514· LMArena data1,403.4
174GPT-5 Chat· LMArena data1,403
175Gemini 3.1 Flash Lite· LMArena data1,402.7
176Grok 4.3· LMArena data1,401.3
177Hunyuan T120250711· LMArena data1,398.7
178GPT-5.3 Chat· LMArena data1,396.9
179Muse Glimmer· LMArena data1,396.8
180Nemotron 3 Super 120B A12B· LMArena data1,396
181GPT 4.12025-04-14· LMArena data1,393.1
182Qwen 3 235B A22BNon-Thinking· LMArena data1,392.2
183GLM 4.6v· LMArena data1,391
184Granite 4.2 30B· LMArena data1,390.6
185GLM 4.5 Air· LMArena data1,387.9
186GPT 5.4 NanoHigh· LMArena data1,387.3
187Ling Flash 2.0· LMArena data1,386.8
188Nova 2 Lite· LMArena data1,385.4
189Claude Sonnet 4Thinking 32K, 20250514· LMArena data1,385.1
190Qwen3 Next 80B A3Bthinking· LMArena data1,384.2
191GPT-5 miniHigh· LMArena data1,383.6
192MiniMax M2.5· LMArena data1,383.5
193Gemini 2.5 Flash-Lite PreviewNon-Thinking, 09-2025· LMArena data1,382.6
194Claude Opus 420250514· LMArena data1,381.7
195Qwen3 Coder· LMArena data1,380.3
196DeepSeek V30324· LMArena data1,378.9
197Step3· LMArena data1,378.3
198Kimi K2 Preview0905· LMArena data1,378
199Grok 3 MiniHigh· LMArena data1,377.6
200Ring Flash 2.0· LMArena data1,377.4

Showing 151–200 of 421 results.

Data credits: LMArena, CC BY 4.0 (licence)