Skip to content

Benchmark

Arena ELO - Exclude Ties

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

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.28

Score distribution

1,247.7
Mean
1,316.3
Median
217.2
Std dev
908.5
10th percentile
1,465.2
90th percentile
533.7
Min
533.71,565.6
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.28 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
251Olmo 3.1 32B Instruct· LMArena data1,250
252Hunyuan Turbo0110· LMArena data1,248.5
253Gemma 3n E4B IT· LMArena data1,245
254Grok 22024-08-13· LMArena data1,243.9
255Yi Lightning· LMArena data1,241.2
256Claude Sonnet 3.720250219· LMArena data1,240.4
257Granite 4.2 3B· LMArena data1,240.1
258GPT 4o2024-05-13· LMArena data1,238.2
259Olmo 3 32B Think· LMArena data1,237.8
260Granite 4.1 8B· LMArena data1,236.3
261Qwen2.5-Plus1127· LMArena data1,236.2
262Claude 3.5 Sonnet20241022· LMArena data1,236
263DeepSeek V2.51210· LMArena data1,228.7
264Athene V2 Chat· LMArena data1,223.6
265GPT OSS 20b· LMArena data1,222.2
266GLM 4 Plus· LMArena data1,221.2
267Gemma 3 4B IT· LMArena data1,221
268Llama 4 Maverick· LMArena data1,220.5
269Molmo 2 8B· LMArena data1,217.6
270Hunyuan Large2025-02-10· LMArena data1,217
271GPT-4o mini2024-07-18· LMArena data1,216
272Gemini 1.5 Flash002· LMArena data1,215.3
273Llama 3.1 405B InstructBF16· LMArena data1,213.4
274Mercury· LMArena data1,211.6
275Llama 3.1 Nemotron 70B Instruct· LMArena data1,211
276GPT-4.1 nano2025-04-14· LMArena data1,210.3
277GPT 4o2024-08-06· LMArena data1,209
278Llama 3.1 405B InstructFP8· LMArena data1,208.9
279Grok 2 mini2024-08-13· LMArena data1,208.3
280Llama 4 Scout· LMArena data1,207.8
281Claude 3.5 Sonnet20240620· LMArena data1,205.9
282Qwen Max0919· LMArena data1,205.8
283Gemini Advanced0514· LMArena data1,203.2
284Mistral Small 3.1 24B Instruct2503· LMArena data1,203.1
285Hunyuan Standard2025-02-10· LMArena data1,199.8
286Llama 3.3 70B Instruct· LMArena data1,199.4
287Gemini 1.5 Pro001· LMArena data1,195.1
288DeepSeek V2.5· LMArena data1,192.7
289GPT-4 Turbo2024-04-09· LMArena data1,190.1
290Olmo 3.1 32B Think· LMArena data1,189.9
291Qwen2.5 72B Instruct· LMArena data1,189.1
292Athene 70B0725· LMArena data1,187.8
293Hunyuan Large Vision· LMArena data1,187.1
294Mistral Large 22407· LMArena data1,185.5
295Mistral Large2411· LMArena data1,179.3
296Magistral Medium2506· LMArena data1,177.7
297Llama 3.1 70B Instruct· LMArena data1,176.6
298Claude 3 Opus20240229· LMArena data1,174.8
299GPT-4 Turbo Preview1106· LMArena data1,174
300GPT-4 Turbo Preview0125· LMArena data1,171.5

Showing 251–300 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)