Skip to content

Benchmark

Arena ELO - Hard Prompts

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

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.24

Score distribution

1,308.5
Mean
1,356.7
Median
162.3
Std dev
1,065.4
10th percentile
1,480.3
90th percentile
727.9
Min
727.91,550.7
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.24 for chat models, measured over 366 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
201DeepSeek V30324· LMArena data1,364.8
202Mistral Medium 32505· LMArena data1,364.3
203Nova 2 Lite· LMArena data1,364
204Grok 3 Minibeta· LMArena data1,363.9
205GPT OSS 120B· LMArena data1,363.2
206Mercury 2· LMArena data1,362.5
207Ling Flash 2.0· LMArena data1,362.5
208Qwen 3 235B A22B· LMArena data1,362.1
209DeepSeek R1· LMArena data1,361.1
210Qwen2.5-Max· LMArena data1,359.5
211MiniMax M2· LMArena data1,356.7
212GLM 4.7 Flash· LMArena data1,355.7
213nova-experimental-chat10-09· LMArena data1,355.4
214Claude Sonnet 420250514· LMArena data1,354.7
215Step3· LMArena data1,354.7
216O1 Preview· LMArena data1,354
217INTELLECT-3· LMArena data1,352
218o4 Mini2025-04-16· LMArena data1,350.2
219Trinity Large Preview· LMArena data1,349.9
220GPT-4.1 mini2025-04-14· LMArena data1,348.8
221Nemotron 3 Nano 30B A3BBF16· LMArena data1,348
222Gemini 2.0 Flash001· LMArena data1,346.1
223Ring Flash 2.0· LMArena data1,344.3
224Trinity Large Thinking· LMArena data1,341.9
225Nemotron 3.5 Lightning 30B A3BNVFP4· LMArena data1,340.5
226Gemma 3 27B IT· LMArena data1,340.4
227MiniMax M1· LMArena data1,339.4
228Llama 3.3 Nemotron Super 49B v1.5· LMArena data1,335.7
229Mistral Small 3.22506· LMArena data1,335.3
230Step-1o Turbo202506· LMArena data1,335
231Qwen3 32B· LMArena data1,334.4
232GLM 4.5V· LMArena data1,334.2
233Claude Sonnet 3.7Thinking 32K, 20250219· LMArena data1,333.2
234O1 Mini· LMArena data1,333.1
235o3 Mini· LMArena data1,332.7
236GPT-5 nanoHigh· LMArena data1,328.2
237Granite 4.2 8B· LMArena data1,326.2
238Command A03-2025· LMArena data1,326.1
239QwQ 32B· LMArena data1,325.4
240Hunyuan TurboS20250226· LMArena data1,324.7
241Gemini 2.0 Flash-Lite Preview02-05· LMArena data1,324.2
242Olmo 3.1 32B Instruct· LMArena data1,322
243Qwen Plus0125· LMArena data1,316.5
244Llama 3.1 Nemotron Ultra 253B v1· LMArena data1,315.9
245Claude Sonnet 3.720250219· LMArena data1,315
246Qwen3 30B A3B· LMArena data1,313.8
247DeepSeek V3· LMArena data1,311.9
248Llama 3.3 Nemotron Super 49B v1· LMArena data1,311.2
249Gemma 3 12B IT· LMArena data1,309.2
250Hunyuan Turbo0110· LMArena data1,308

Showing 201–250 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)