Skip to content

Benchmark

Arena ELO - English

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

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.26

Score distribution

1,327.7
Mean
1,373.1
Median
138.5
Std dev
1,119.5
10th percentile
1,468.7
90th percentile
845.9
Min
845.91,532
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.26 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
351Reka Flash 21B20240226· LMArena data1,179.1
352Zephyr ORPO 141B A35B v0.1· LMArena data1,174.4
353Granite 3.1 2B Instruct· LMArena data1,172.5
354Phi-3 Medium 4K Instruct· LMArena data1,169.3
355Mixtral 8x7B Instruct v0.1· LMArena data1,167.6
356Starling LM 7B Beta· LMArena data1,161.9
357Gemini ProDev API· LMArena data1,161.9
358Qwen1.5-32B-Chat· LMArena data1,159.9
359Yi 34B Chat· LMArena data1,159.1
360Llama 2 70B Chat· LMArena data1,156
361DBRX Instruct Preview· LMArena data1,155.1
362Gemini Pro· LMArena data1,153.7
363Phi-3 Small 8K Instruct· LMArena data1,150.6
364WizardLM 70B· LMArena data1,150.4
365Tulu 2 DPO 70B· LMArena data1,150
366Granite 3.0 8B Instruct· LMArena data1,149.3
367Qwen1.5-14B-Chat· LMArena data1,148
368Nous Hermes 2 Mixtral 8x7B DPO· LMArena data1,139.6
369GPT-3.5 Turbo0125· LMArena data1,139.2
370Starling LM 7B Alpha· LMArena data1,138.7
371Vicuna 33B· LMArena data1,133.5
372Mistral 7B Instruct v0.2· LMArena data1,133.5
373OpenChat 3.50106· LMArena data1,131.2
374Phi-3 Mini 4K InstructJune 2024· LMArena data1,128.8
375DeepSeek LLM 67B Chat· LMArena data1,127.1
376OpenHermes 2.5 Mistral 7B· LMArena data1,125.6
377Gemma 1.1 7B IT· LMArena data1,124.1
378Snowflake Arctic Instruct· LMArena data1,121.6
379Llama 2 13B Chat· LMArena data1,119.9
380Llama 2 70B SteerLM Chat· LMArena data1,119.8
381Granite 3.0 2B Instruct· LMArena data1,119.5
382OpenChat 3.5· LMArena data1,117.7
383SOLAR 10.7B Instruct v1.0· LMArena data1,117.3
384Phi-3 Mini 4K Instruct· LMArena data1,117.3
385Llama 3.2 1B Instruct· LMArena data1,116.7
386GPT-3.5 Turbo1106· LMArena data1,113.1
387Dolphin 2.2.1 Mistral 7B· LMArena data1,112.5
388Zephyr 7B Beta· LMArena data1,107.4
389Qwen1.5-7B-Chat· LMArena data1,105.3
390MPT 30B Chat· LMArena data1,102.2
391WizardLM 13B· LMArena data1,101.7
392Code Llama 70B Instruct· LMArena data1,095.9
393Llama 2 7B Chat· LMArena data1,095.9
394SmolLM2 1.7B Instruct· LMArena data1,092.7
395Code Llama 34B Instruct· LMArena data1,092
396Zephyr 7B Alpha· LMArena data1,091.7
397Gemma 7B IT· LMArena data1,090.1
398Guanaco 33B· LMArena data1,085.4
399Phi-3 Mini 128K Instruct· LMArena data1,083
400Falcon 180B Chat· LMArena data1,080.7

Showing 351–400 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)