Skip to content

Benchmark

Arena ELO - Industry Software and It Services

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

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.29

Score distribution

1,322.6
Mean
1,374.8
Median
161.3
Std dev
1,081.5
10th percentile
1,486.1
90th percentile
719.3
Min
719.31,552.3
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.29 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
351Starling LM 7B Beta· LMArena data1,141.5
352Qwen1.5-14B-Chat· LMArena data1,140.4
353Granite 3.1 2B Instruct· LMArena data1,139.5
354Zephyr ORPO 141B A35B v0.1· LMArena data1,133.6
355Phi-3 Medium 4K Instruct· LMArena data1,133.3
356Gemma 2 2B IT· LMArena data1,132.1
357Gemini ProDev API· LMArena data1,131.1
358Mixtral 8x7B Instruct v0.1· LMArena data1,124.9
359Yi 34B Chat· LMArena data1,121.8
360DBRX Instruct Preview· LMArena data1,116.5
361GPT-3.5 Turbo0125· LMArena data1,115.3
362Gemini Pro· LMArena data1,111
363Tulu 2 DPO 70B· LMArena data1,110.1
364Phi-3 Small 8K Instruct· LMArena data1,108.4
365OpenChat 3.50106· LMArena data1,104.8
366Llama 3.2 3B· LMArena data1,103.2
367Granite 3.0 8B Instruct· LMArena data1,100
368Llama 2 70B Chat· LMArena data1,098.9
369Gemma 1.1 7B IT· LMArena data1,095.1
370DeepSeek LLM 67B Chat· LMArena data1,094.9
371Nous Hermes 2 Mixtral 8x7B DPO· LMArena data1,094.7
372Starling LM 7B Alpha· LMArena data1,091.3
373GPT-3.5 Turbo1106· LMArena data1,091.2
374Qwen1.5-7B-Chat· LMArena data1,090.1
375WizardLM 70B· LMArena data1,088.4
376Phi-3 Mini 4K InstructJune 2024· LMArena data1,088
377Snowflake Arctic Instruct· LMArena data1,087.4
378Vicuna 33B· LMArena data1,085.8
379OpenChat 3.5· LMArena data1,083.3
380Mistral 7B Instruct v0.2· LMArena data1,082.8
381Phi-3 Mini 4K Instruct· LMArena data1,081.5
382Granite 3.0 2B Instruct· LMArena data1,079
383OpenHermes 2.5 Mistral 7B· LMArena data1,076.8
384Llama 2 13B Chat· LMArena data1,072.7
385SOLAR 10.7B Instruct v1.0· LMArena data1,067.5
386Code Llama 34B Instruct· LMArena data1,062.9
387Gemma 7B IT· LMArena data1,059.7
388MPT 30B Chat· LMArena data1,057.6
389Code Llama 70B Instruct· LMArena data1,057.6
390Llama 3.2 1B Instruct· LMArena data1,056.2
391Llama 2 70B SteerLM Chat· LMArena data1,055.7
392Zephyr 7B Alpha· LMArena data1,053.2
393Qwen-14B-Chat· LMArena data1,052.8
394WizardLM 13B· LMArena data1,052.2
395Zephyr 7B Beta· LMArena data1,048.2
396Vicuna 13B· LMArena data1,047.8
397SmolLM2 1.7B Instruct· LMArena data1,046.6
398Dolphin 2.2.1 Mistral 7B· LMArena data1,046.5
399OLMo 7B Instruct· LMArena data1,037.1
400Phi-3 Mini 128K Instruct· LMArena data1,033.7

Showing 351–400 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)