Skip to content

Benchmark

Arena ELO - Hard Prompts

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

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.24

Score distribution

1,308.5
Mean
1,356.7
Median
162.3
Std dev
1,065.4
10th percentile
1,480.3
90th percentile
727.9
Min
727.91,550.7
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.24 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
1Gemini 4 CarbonHigh· LMArena data1,550.7
2Gemini 4 NeonHigh· LMArena data1,547.1
3Gemini 4 ArgonHigh· LMArena data1,541.7
4Claude Opus 5.5High· LMArena data1,530.9
5Claude Fable 5.1Max· LMArena data1,527.3
6Claude Opus 5Max· LMArena data1,525.2
7Claude Opus 4.6High· LMArena data1,524.4
8Claude Opus 4.6· LMArena data1,522.2
9Claude Opus 5High· LMArena data1,522
10Gemini 4 HydrogenHigh· LMArena data1,517.2
11Muse WonderMax· LMArena data1,513.6
12MiMo V2.6 Pro· LMArena data1,511.7
13Gemini 3.8 FlashHigh· LMArena data1,509.2
14Claude Fable 5High· LMArena data1,507.7
15Claude Opus 4.7High· LMArena data1,505.8
16Muse Spark 1.3Max· LMArena data1,504.2
17Qwen3.8 Max· LMArena data1,500.3
18Claude Opus 4.7· LMArena data1,499.9
19Gemini 3.7 FlashHigh· LMArena data1,498.4
20Kimi K3Max· LMArena data1,494.1
21Muse Spark 1.2xHigh· LMArena data1,492.6
22GLM 5.3Max· LMArena data1,492.5
23GLM 5.3 Flash· LMArena data1,490.4
24Gemini 3.5 FlashHigh· LMArena data1,489.3
25GPT 5.5High· LMArena data1,488.3
26Claude Sonnet 5.5xHigh· LMArena data1,487.2
27Muse Spark 1.1· LMArena data1,486.2
28Gemini 3.6 FlashHigh· LMArena data1,485.7
29GPT 5.5· LMArena data1,485.5
30MiMo V2.5 Pro· LMArena data1,485.4
31Gemini 3.5 FlashMedium· LMArena data1,485.3
32GPT 5.4High· LMArena data1,484.9
33Gemini 3.1 Pro Preview· LMArena data1,484.5
34Claude Sonnet 4.6· LMArena data1,484.4
35Claude Fable 5.2Max· LMArena data1,484.3
36Qwen3.5-Max-Preview· LMArena data1,484
37GPT-5.6 SolxHigh· LMArena data1,483
38Claude Opus 4.8High· LMArena data1,482.3
39DeepSeek V4.1 Flash 0910Max· LMArena data1,481.8
40Qwen3.7 Max Preview· LMArena data1,481.2
41ERNIE 5.1· LMArena data1,480.8
42GLM 5.2Max· LMArena data1,480.6
43Gemini 3 Pro Preview· LMArena data1,480.3
44Claude Opus 4.520251101· LMArena data1,476.4
45Muse Spark· LMArena data1,474.4
46Claude Opus 4.8· LMArena data1,473.4
47Claude Opus 4.5High, 32K, 20251101· LMArena data1,473.3
48GLM 5.1· LMArena data1,472.7
49Kimi K2.6· LMArena data1,470.4
50GPT-5.6 TerraxHigh· LMArena data1,469.8

Showing 1–50 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)