Skip to content

Benchmark

Design Arena Elo (agents/agon_slides)

Agentic benchmarks measure how well models carry out tasks that take several steps. Higher scores are better.

Models measured
31
Scores recorded
32
Sources
1
IQ weight
1.44

Score distribution

1,161.3
Mean
1,161
Median
87.2
Std dev
1,046
10th percentile
1,269
90th percentile
982
Min
9821,378
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.44 for chat models, measured over 31 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Claude Opus 5.5high· Design Arena Leaderboards data1,378
2Claude Fable 5.1· Design Arena Leaderboards data1,318
3Kimi K3.1 Pro· Design Arena Leaderboards data1,287
4Claude Fable 5· Design Arena Leaderboards data1,269
5Qwen3.8 Max· Design Arena Leaderboards data1,243
6Claude Opus 5· Design Arena Leaderboards data1,225
7Kimi K3· Design Arena Leaderboards data1,223
8GLM 5.3· Design Arena Leaderboards data1,217
9Muse Spark 1.3· Design Arena Leaderboards data1,211
10GPT-6 Astra· Design Arena Leaderboards data1,207
11GPT-5.6 Sol· Design Arena Leaderboards data1,205
12Claude Opus 4.8· Design Arena Leaderboards data1,188
13GPT-5.6 Solxhigh· Design Arena Leaderboards data1,185
14Claude Sonnet 5· Design Arena Leaderboards data1,179
15MiniMax M3· Design Arena Leaderboards data1,167
16GLM 5.2· Design Arena Leaderboards data1,161
17GLM 5.1· Design Arena Leaderboards data1,157
18Muse Spark 1.2· Design Arena Leaderboards data1,157
19Qwen3.7 Max· Design Arena Leaderboards data1,151
20Gemini 3.5 Flash· Design Arena Leaderboards data1,150
21Gemini 3.8 Flash· Design Arena Leaderboards data1,149
22Kimi K2.7 Code· Design Arena Leaderboards data1,119
23GLM 5V Turbo· Design Arena Leaderboards data1,115
24Gemini 3.6 Flash· Design Arena Leaderboards data1,106
25Kimi K2.6· Design Arena Leaderboards data1,096
26GPT-5.6 Terra· Design Arena Leaderboards data1,092
27GPT 5.5· Design Arena Leaderboards data1,078
28Gemini 3.7 Flash· Design Arena Leaderboards data1,052
29Gemini 3.1 Pro Preview· Design Arena Leaderboards data1,046
30Muse Wonder 1.0· Design Arena Leaderboards data1,033
31Gemini 3 Pro Preview· Design Arena Leaderboards data1,015
32Grok 4.3· Design Arena Leaderboards data982