Skip to content

Benchmark

Arena ELO - Polish

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

Models measured
207
Scores recorded
242
Sources
1
IQ weight
2.07

Score distribution

1,386.8
Mean
1,405.7
Median
84.1
Std dev
1,270
10th percentile
1,481.7
90th percentile
1,072
Min
1,0721,546.7
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.07 for chat models, measured over 207 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
201GPT-4 Turbo2024-04-09· LMArena data1,301.8
202Step3· LMArena data1,300.6
203Solar Pro 4· LMArena data1,300.4
204GPT-4 Turbo Preview0125· LMArena data1,299.1
205Qwen3 30B A3B· LMArena data1,299
206o3 Mini· LMArena data1,294.2
207Gemini 1.5 Pro001· LMArena data1,291
208Trinity Large Preview· LMArena data1,289.1
209GLM 4.7 Flash· LMArena data1,288.7
210QwQ 32B· LMArena data1,287.9
211Claude 3.5 Sonnet20241022· LMArena data1,287.8
212Gemma 3n E4B IT· LMArena data1,285.4
213Claude Sonnet 3.7Thinking 32K, 20250219· LMArena data1,284.4
214Claude Sonnet 3.720250219· LMArena data1,278.4
215Gemma 2 27B IT· LMArena data1,277.2
216Llama 4 Maverick· LMArena data1,275.3
217Ring Flash 2.0· LMArena data1,273.9
218Nemotron 3.5 Lightning 30B A3BNVFP4· LMArena data1,270
219GPT 40314· LMArena data1,265.1
220Llama 4 Scout· LMArena data1,264.7
221Claude 3 Sonnet20240229· LMArena data1,259.3
222Llama 3.3 70B Instruct· LMArena data1,259.2
223GPT OSS 20b· LMArena data1,252.5
224Olmo 3 32B Think· LMArena data1,251.6
225Claude Haiku 3.520241022· LMArena data1,249.4
226Magistral Medium2506· LMArena data1,248.1
227GPT-4o mini2024-07-18· LMArena data1,243.9
228Mistral Small 3.1 24B Instruct2503· LMArena data1,241.8
229Command R+· LMArena data1,238.3
230Gemini 1.5 Flash001· LMArena data1,237.6
231Hunyuan Large Vision· LMArena data1,236.4
232Claude 3 Haiku20240307· LMArena data1,233.4
233Olmo 3.1 32B Instruct· LMArena data1,220.8
234GPT 40613· LMArena data1,219.4
235Llama 3 70B Instruct· LMArena data1,196.3
236Olmo 3.1 32B Think· LMArena data1,191.4
237Mixtral 8x22B Instruct v0.1· LMArena data1,170.9
238Granite 4.0 H Small· LMArena data1,168.1
239Mistral Large2402· LMArena data1,157.1
240Llama 3 8B Instruct· LMArena data1,136.9
241GPT-3.5 Turbo0125· LMArena data1,121.2
242Mixtral 8x7B Instruct v0.1· LMArena data1,072

Showing 201–242 of 242 results.

Data credits: LMArena, CC BY 4.0 (licence)