The Agents Leaderboard ranks large language models by their ability to act as autonomous problem-solvers rather than simply generate responses. Unlike standard text-generation evaluations, agentic benchmarks test whether a model can independently break down complex goals, execute shell commands in terminal environments, invoke external APIs, interpret tool outputs, and recover from runtime errors.

Drawing on evaluations agent-focused benchmarks this leaderboard highlights models capable of handling complex, multi-step tasks across developer and enterprise workflows.

Looking for well-rounded performance? Agentic capabilities account for 20% of the composite evaluation. Check the Overall LLM Ranking to see how top agentic models perform across coding, mathematics, and complex reasoning.

Last updated: November 5, 2026
Rank Model Name Relative Quality Score
1 Claude Opus 5.5
Quality of Agents 97.90%

97.90
2 Gemini 4 Carbon
Quality of Agents 95.55%

95.55
3 GPT-6 Astra
Quality of Agents 94.17%

94.17
4 Claude Fable 5.1
Quality of Agents 90.37%

90.37
5 Claude Sonnet 5.5
Quality of Agents 88.73%

88.73
6 GPT-6.1 Sol
Quality of Agents 85.37%

85.37
7 Muse Wonder Full
Quality of Agents 82.79%

82.79
8 Gemini 4 Argon
Quality of Agents 82.30%

82.30
9 Claude Opus 4.8
Quality of Agents 79.45%

79.45
10 Claude Fable 5
Quality of Agents 78.45%

78.45
11 Claude Opus 5
Quality of Agents 77.85%

77.85
12 GPT-6.1 Terra
Quality of Agents 76.63%

76.63
13 GPT-5.5
Quality of Agents 75.17%

75.17
14 GPT-5.6 Sol
Quality of Agents 74.30%

74.30
15 GPT-6 Sol
Quality of Agents 74.22%

74.22
16 GPT-5.5 Pro
Quality of Agents 73.80%

73.80
17 Claude Opus 4.7
Quality of Agents 72.95%

72.95
18 Grok 4.5
Quality of Agents 72.95%

72.95
19 GPT-5.6 Luna
Quality of Agents 72.77%

72.77
20 DeepSeek-V4-Flash
Quality of Agents 72.60%

72.60
21 Hy4
Quality of Agents 72.10%

72.10
22 Gemini 3.5 Flash
Quality of Agents 70.38%

70.38
23 Beam 1.0
Quality of Agents 70.21%

70.21
24 Ling 3.0 Flash Fin
Quality of Agents 70.21%

70.21
25 Muse Wonder
Quality of Agents 70.11%

70.11
26 Kimi K3.1 Pro
Quality of Agents 68.86%

68.86
27 Claude Opus 4.6
Quality of Agents 68.32%

68.32
28 GLM-5.3
Quality of Agents 68.30%

68.30
29 Kimi K2.6
Quality of Agents 67.47%

67.47
30 Gemini 3.6 Flash
Quality of Agents 66.78%

66.78
31 GPT-5.4
Quality of Agents 66.27%

66.27
32 Kimi K2.7 Code
Quality of Agents 66.27%

66.27
33 Grok 4.7
Quality of Agents 65.16%

65.16
34 DeepSeek-V4-Pro
Quality of Agents 64.93%

64.93
35 Claude Sonnet 5
Quality of Agents 64.53%

64.53
36 DeepSeek-V4.1-Flash
Quality of Agents 63.54%

63.54
37 MiMo-V2.6-Flash
Quality of Agents 63.40%

63.40
38 GLM-5.1
Quality of Agents 62.33%

62.33
39 MiMo-V2.6-Pro
Quality of Agents 62.18%

62.18
40 Claude Sonnet 4.6
Quality of Agents 61.99%

61.99
41 K2 Horizon 375B-A23B
Quality of Agents 61.82%

61.82
42 GLM-5.2
Quality of Agents 61.41%

61.41
43 Claude Haiku 5.5
Quality of Agents 60.92%

60.92
44 Gemini 3.7 Flash
Quality of Agents 60.67%

60.67
45 Muse Spark 1.3
Quality of Agents 58.99%

58.99
46 Qwen3.8-Flash-Next
Quality of Agents 58.00%

58.00
47 Qwen3.8
Quality of Agents 57.87%

57.87
48 Kimi K3
Quality of Agents 57.54%

57.54
49 GPT-5.6 Terra
Quality of Agents 57.22%

57.22
50 MiMo-V2.5
Quality of Agents 56.85%

56.85
51 GPT-6 Luna
Quality of Agents 56.59%

56.59
52 Qwen3.6 Plus
Quality of Agents 56.34%

56.34
53 Grok 4.6
Quality of Agents 56.12%

56.12
54 Qwen3.5-397B-A17B
Quality of Agents 55.14%

55.14
55 MiniMax M3.1 Flash
Quality of Agents 54.51%

54.51
56 GLM-5
Quality of Agents 54.45%

54.45
57 GPT-5.4 mini
Quality of Agents 54.11%

54.11
58 GPT-5.2 Codex
Quality of Agents 54.11%

54.11
59 Ling 3.0 Flash
Quality of Agents 53.94%

53.94
60 Muse Spark 1.1
Quality of Agents 53.65%

53.65
61 Gemini 3.8 Flash
Quality of Agents 53.46%

53.46
62 GLM-5V-Turbo
Quality of Agents 52.91%

52.91
63 Kimi K2.5
Quality of Agents 52.57%

52.57
64 GPT-5.2 Pro
Quality of Agents 52.23%

52.23
65 Qwen3.7
Quality of Agents 51.47%

51.47
66 Gemini 3 Flash
Quality of Agents 51.37%

51.37
67 Claude Opus 4.5
Quality of Agents 51.20%

51.20
68 Mistral Large 4
Quality of Agents 51.06%

51.06
69 MiniMax M2.1
Quality of Agents 51.03%

51.03
70 Gemini 3.1 Pro
Quality of Agents 50.99%

50.99
71 MiniMax M2.5
Quality of Agents 50.68%

50.68
72 Seed 2.0 Pro
Quality of Agents 50.51%

50.51
73 MiMo-V2-Pro
Quality of Agents 50.34%

50.34
74 Claude Sonnet 4.5
Quality of Agents 50.34%

50.34
75 GLM-5.3-Flash
Quality of Agents 50.23%

50.23
76 Hy3
Quality of Agents 48.74%

48.74
77 Qwen3.6-27B
Quality of Agents 48.46%

48.46
78 Gemini 3.5 Flash-Lite
Quality of Agents 48.29%

48.29
79 MiniMax M2.7
Quality of Agents 48.12%

48.12
80 Muse Spark
Quality of Agents 47.60%

47.60
81 Qwen3.5-122B-A10B
Quality of Agents 47.60%

47.60
82 Kimi K2
Quality of Agents 47.43%

47.43
83 Muse Spark 1.2
Quality of Agents 47.21%

47.21
84 Gemini 3 Pro
Quality of Agents 47.09%

47.09
85 Step-3.5-Flash
Quality of Agents 46.92%

46.92
86 Qwen3.7-Plus
Quality of Agents 46.83%

46.83
87 Qwen3.8-27B
Quality of Agents 46.40%

46.40
88 K2 Horizon MoVA 36B-A4B
Quality of Agents 46.40%

46.40
89 GPT-5.2
Quality of Agents 45.89%

45.89
90 Qwen3.5-27B
Quality of Agents 45.89%

45.89
91 Muse Spark Open
Quality of Agents 44.75%

44.75
92 MiniMax M3
Quality of Agents 44.64%

44.64
93 MiMo-V2.5-Pro
Quality of Agents 44.63%

44.63
94 Step 3.7 Flash
Quality of Agents 44.52%

44.52
95 Claude Haiku 4.5
Quality of Agents 43.84%

43.84
96 Qwen3 VL 235B A22B
Quality of Agents 43.49%

43.49
97 LongCat-Flash
Quality of Agents 43.32%

43.32
98 Claude Opus 4.1
Quality of Agents 42.64%

42.64
99 Qwen3.5-35B-A3B
Quality of Agents 42.64%

42.64
100 Gemma 4 26B-A4B
Quality of Agents 42.47%

42.47