Skip to content

Benchmark

Arena Agent

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

Models measured
62
Scores recorded
68
Sources
1
IQ weight
2.67

Score distribution

0.007
Mean
0.006
Median
0.101
Std dev
-0.109
10th percentile
0.137
90th percentile
-0.239
Min
-0.2390.274
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.67 for chat models, measured over 62 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 Fable 5.2Max· LMArena data0.274
2Claude Opus 5.5Max· LMArena data0.186
3Gemini 4 NeonHigh· LMArena data0.184
4Gemini 4 CarbonHigh· LMArena data0.161
5Claude Sonnet 5.5Max· LMArena data0.156
6GPT-6.2 AstraMax· LMArena data0.152
7GPT-6 AstraMax· LMArena data0.146
8Claude Opus 5.5High· LMArena data0.137
9Claude Fable 5.1Max· LMArena data0.137
10Gemini 4 HydrogenHigh· LMArena data0.135
11GPT-6.2 SolMax· LMArena data0.113
12GPT-6.1 SolMax· LMArena data0.113
13GPT-6.1 TerraMax· LMArena data0.105
14Gemini 4 ArgonHigh· LMArena data0.103
15GPT-6.1 SolMedium· LMArena data0.082
16Muse Wonder FullMax· LMArena data0.08
17Claude Sonnet 5.5High· LMArena data0.0742
18Qwen4 MaxxHigh· LMArena data0.0727
19GPT-6 SolMax· LMArena data0.0647
20Claude Opus 5Max· LMArena data0.0626
21Hy 4· LMArena data0.0609
22Claude Fable 5High· LMArena data0.0586
23GPT-6.1 TerraMedium· LMArena data0.0461
24Kimi K3.1 ProMax· LMArena data0.0388
25GPT-5.6 SolxHigh· LMArena data0.0368
26Inkling 2 Small· LMArena data0.0151
27Gemini 3.8 FlashHigh· LMArena data0.0149
28Grok 4.7xHigh· LMArena data0.0148
29MiMo V2.6 Pro· LMArena data0.0142
30Muse Spark 1.3Max· LMArena data0.0111
31Claude Sonnet 5High· LMArena data0.01
32Muse EmberMax· LMArena data0.00965
33DeepSeek V4.1 Flash 1017Max· LMArena data0.00785
34GPT-6 LunaMax· LMArena data0.0064
35Muse WonderMax· LMArena data0.00583
36GLM 5.3Max· LMArena data-0.000421
37Kimi K3Max· LMArena data-0.00148
38GLM 5.2Max· LMArena data-0.0041
39Hy4 preview· LMArena data-0.00439
40Grok 4.6xHigh· LMArena data-0.00596
41Claude Haiku 5.5High· LMArena data-0.00613
42Qwen3.8 Max· LMArena data-0.00913
43MiMo V2.6 Flash· LMArena data-0.0175
44GPT-5.6 TerraxHigh· LMArena data-0.0267
45GPT-6 LunaMedium· LMArena data-0.0305
46Gemini 3.7 FlashHigh· LMArena data-0.0307
47MiniMax M3.1 Flash· LMArena data-0.0367
48GLM 5.3 Flash· LMArena data-0.0387
49Qwen 3.8 Flash Next· LMArena data-0.0491
50Step 5 Preview· LMArena data-0.0515

Showing 1–50 of 68 results.

Data credits: LMArena, CC BY 4.0 (licence)