MMODELYST
CAP

Agents

League table — best at agents · 30
mean percentile across this capability's benchmarks
#ModelLeagueCoverage$/1Mtok/s
1Claude Fable 598.92/3$2065
2GLM-5.2 (max)97.82/3$2.15200
3Gemini 3.1 Pro Preview95.92/3$4.5140
4Claude Opus 4.895.02/3$1059
5DeepSeek V4 Pro95.02/3$0.54464
6Qwen3.6 Plus94.82/3$1.1353
7GLM-594.62/3$1.5551
8GPT-5.594.42/3$11.2572
9Qwen3.6 Max Preview94.22/3$2.9253
10GLM-5.194.22/3$2.1575
11Qwen3.7 Max94.12/3$3.75197
12Kimi K2.693.42/3$1.7143
13Grok 4.20 030992.62/3$3
14Qwen3.5 397B A17B91.52/3$1.3562
15Grok 4.391.52/3$1.56152
16Qwen3.7 Plus91.12/3$0.752
17MiMo-V2-Pro91.02/3$1.5
18Step 3.7 Flash91.02/3$0.438404
19MiMo-V2.5-Pro90.92/3$0.54455
20Gemini 3.5 Flash90.72/3$3.38238
21Claude Opus 4.690.22/3$1048
22Muse Spark89.22/3
23GLM-5-Turbo89.22/3
24Kimi K2.589.12/3$1.249
25KAT Coder Pro V289.02/3$0.525
26Claude Opus 4.588.92/3$1057
27Claude Opus 4.788.82/3$1052
28GLM 5V Turbo88.72/3
29GPT-5.488.72/3$5.63161
30MiniMax-M2.588.32/3$0.52578

League = a model's mean percentile across the 3 benchmarksin this capability (best score per benchmark; partial coverage shown). Price & speed via Artificial Analysis.

Benchmarks in this capability (7)
ALFWorld0 scores
BabyAI0 scores
SWE-Gym0 scores
WebArena0 scores
τ²-bench324 scores