Skip to content

Benchmark

Arena ELO - Industry Writing and Literature and Language

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

Models measured
365
Scores recorded
422
Sources
1
IQ weight
1.17

Score distribution

1,297.8
Mean
1,328.7
Median
142.8
Std dev
1,086
10th percentile
1,453.9
90th percentile
809.1
Min
809.11,540.8
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.17 for chat models, measured over 365 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
251Qwen3 30B A3B· LMArena data1,290
252GPT 4o2024-08-06· LMArena data1,289.7
253Gemini 1.5 Pro001· LMArena data1,288.8
254Gemma 3 4B IT· LMArena data1,288
255Gemma 3n E4B IT· LMArena data1,287
256Gemini Advanced0514· LMArena data1,286.2
257Nemotron 3 Nano 30B A3BBF16· LMArena data1,283.4
258GPT-5 nanoHigh· LMArena data1,283.3
259DeepSeek V2.51210· LMArena data1,283.1
260Nemotron 3.5 Lightning 30B A3BNVFP4· LMArena data1,283
261GPT-4o mini2024-07-18· LMArena data1,282
262Claude 3.5 Sonnet20240620· LMArena data1,280.1
263O1 Mini· LMArena data1,280
264GPT-4 Turbo2024-04-09· LMArena data1,278.7
265GLM 4 Plus· LMArena data1,278.6
266Llama 4 Maverick· LMArena data1,277.3
267Yi Lightning· LMArena data1,276.3
268Olmo 3.1 32B Instruct· LMArena data1,275.2
269Llama 3.1 405B InstructFP8· LMArena data1,274.2
270Qwen2.5-Plus1127· LMArena data1,273.5
271Hunyuan Large2025-02-10· LMArena data1,271
272Llama 3.1 405B InstructBF16· LMArena data1,270.5
273Qwen Max0919· LMArena data1,270.3
274Claude 3 Opus20240229· LMArena data1,268.6
275Mistral Small 3.1 24B Instruct2503· LMArena data1,268.4
276Llama 4 Scout· LMArena data1,268.3
277Granite 4.2 8B· LMArena data1,268.1
278Olmo 3 32B Think· LMArena data1,267.6
279GPT-4 Turbo Preview1106· LMArena data1,264.4
280Grok 2 mini2024-08-13· LMArena data1,262.2
281Granite 4.1 8B· LMArena data1,261.9
282GPT-4.1 nano2025-04-14· LMArena data1,260.9
283Mistral Large2411· LMArena data1,260.5
284Hunyuan Large Vision· LMArena data1,259.1
285Mistral Large 22407· LMArena data1,257.9
286GPT-4 Turbo Preview0125· LMArena data1,257.2
287Hunyuan Standard2025-02-10· LMArena data1,257.1
288Magistral Medium2506· LMArena data1,256.6
289Claude Haiku 3.520241022· LMArena data1,256.3
290Llama 3.3 70B Instruct· LMArena data1,255
291Llama 3.1 Tulu 3 70B· LMArena data1,254.3
292Athene V2 Chat· LMArena data1,254.1
293Llama 3.1 Nemotron 70B Instruct· LMArena data1,252.6
294Gemini 1.5 Flash001· LMArena data1,251.8
295Olmo 3.1 32B Think· LMArena data1,250.5
296Gemma 2 27B IT· LMArena data1,248.7
297Qwen2.5 72B Instruct· LMArena data1,246.9
298DeepSeek V2.5· LMArena data1,246.9
299Nova Pro 1.0· LMArena data1,246.6
300Reka Core20240904· LMArena data1,242.4

Showing 251–300 of 422 results.

Data credits: LMArena, CC BY 4.0 (licence)