Skip to content

Benchmark

Arena ELO - Korean

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

Models measured
244
Scores recorded
288
Sources
1
IQ weight
1.99

Score distribution

1,298.9
Mean
1,348.8
Median
139.4
Std dev
1,087.1
10th percentile
1,441.2
90th percentile
856
Min
8561,509.5
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.99 for chat models, measured over 244 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
251Gemini 1.5 Flash-8B001· LMArena data1,140.1
252Command R+· LMArena data1,137.9
253Gemma 2 9B IT· LMArena data1,136.7
254Claude 3 Sonnet20240229· LMArena data1,127.8
255Magistral Medium2506· LMArena data1,124.8
256Claude 3 Haiku20240307· LMArena data1,108.5
257Reka Flash 21BOnline, 20240226· LMArena data1,104.1
258DeepSeek Coder V2 Instruct· LMArena data1,104
259GPT 40314· LMArena data1,087.1
260Qwen2 72B Instruct· LMArena data1,082.7
261Reka Flash 21B20240226· LMArena data1,068.4
262GPT 40613· LMArena data1,059.8
263Mixtral 8x22B Instruct v0.1· LMArena data1,056.7
264Gemma 2 2B IT· LMArena data1,054.8
265Llama 3.1 8B Instruct· LMArena data1,052.8
266Qwen1.5-72B-Chat· LMArena data1,049.5
267Qwen1.5-110B-Chat· LMArena data1,043.7
268GLM 40520· LMArena data1,042.1
269Mistral Medium· LMArena data1,032.1
270GPT-3.5 Turbo0125· LMArena data1,018.8
271Mistral Large2402· LMArena data1,017.1
272Llama 3 70B Instruct· LMArena data1,016.7
273Qwen1.5-32B-Chat· LMArena data1,007.6
274Yi 1.5 34B Chat· LMArena data1,005.3
275Llama 3 8B Instruct· LMArena data1,003.6
276DBRX Instruct Preview· LMArena data992.9
277Gemma 1.1 7B IT· LMArena data987.9
278Mixtral 8x7B Instruct v0.1· LMArena data967.9
279Llama 2 70B Chat· LMArena data964
280Yi 34B Chat· LMArena data959.4
281Phi-3 Medium 4K Instruct· LMArena data954
282Llama 2 13B Chat· LMArena data953.4
283Snowflake Arctic Instruct· LMArena data939.6
284Phi-3 Mini 4K Instruct· LMArena data905.5
285Gemma 1.1 2B IT· LMArena data899
286Phi-3 Small 8K Instruct· LMArena data893.9
287Phi-3 Mini 4K InstructJune 2024· LMArena data879.8
288Phi-3 Mini 128K Instruct· LMArena data856

Showing 251–288 of 288 results.

Data credits: LMArena, CC BY 4.0 (licence)