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
151DeepSeek V3.1Thinking· LMArena data1,336.8
152Grok 3 MiniHigh· LMArena data1,335
153Ling Flash 2.0· LMArena data1,334.4
154Grok 4 FastNon-Reasoning· LMArena data1,331.6
155Qwen3 30B A3B Instruct2507· LMArena data1,330.9
156DeepSeek R1· LMArena data1,329.5
157Kimi K2 Thinking· LMArena data1,328.7
158Gemini 2.0 Flash-Lite Preview02-05· LMArena data1,325.2
159nova-experimental-chat10-20· LMArena data1,323.2
160Kimi K2 Preview0905· LMArena data1,322.6
161DeepSeek V30324· LMArena data1,320.3
162Claude Opus 4Thinking 16K, 20250514· LMArena data1,320.1
163Qwen 3 235B A22B· LMArena data1,317.5
164GPT 5.4 NanoHigh· LMArena data1,316.8
165MiniMax M2.7· LMArena data1,313.7
166o3 MiniHigh· LMArena data1,313.6
167Gemini 2.0 Flash001· LMArena data1,312.9
168o4 Mini2025-04-16· LMArena data1,312
169Qwen3 Next 80B A3Bthinking· LMArena data1,308.3
170GLM 4.5 Air· LMArena data1,307.7
171Gemma 3 27B IT· LMArena data1,307
172Trinity Large Thinking· LMArena data1,305.8
173Qwen3 Coder· LMArena data1,305.6
174Grok 3 Minibeta· LMArena data1,304.9
175GPT-5 miniHigh· LMArena data1,304.4
176Mistral Medium 32505· LMArena data1,303.7
177Qwen2.5-Max· LMArena data1,303.6
178Claude Opus 420250514· LMArena data1,303.5
179GPT-4.1 mini2025-04-14· LMArena data1,298
180Gemini 1.5 Pro002· LMArena data1,297.9
181MiniMax-M2.1 Preview· LMArena data1,293.4
182Claude Sonnet 420250514· LMArena data1,291.7
183O1 Preview· LMArena data1,290.5
184GLM 4.7 Flash· LMArena data1,289.7
185Kimi K2 Preview0711· LMArena data1,286.7
186Command A03-2025· LMArena data1,285
187MiMo V2 FlashThinking· LMArena data1,282.9
188GPT OSS 120B· LMArena data1,282
189Nova 2 Lite· LMArena data1,278.9
190QwQ 32B· LMArena data1,278.5
191Claude Sonnet 4Thinking 32K, 20250514· LMArena data1,271.7
192Nemotron 3 Nano 30B A3BBF16· LMArena data1,269.9
193Step3· LMArena data1,269.3
194GPT-5 nanoHigh· LMArena data1,269
195Ring Flash 2.0· LMArena data1,265.1
196MiniMax M1· LMArena data1,264.9
197Qwen3 30B A3B· LMArena data1,261
198Gemma 3n E4B IT· LMArena data1,259.9
199o3 Mini· LMArena data1,259.4
200Trinity Large Preview· LMArena data1,255.5

Showing 151–200 of 288 results.

Data credits: LMArena, CC BY 4.0 (licence)