The Math Ranking evaluates the mathematical rigor and quantitative reasoning capabilities of large language models. Rather than focusing on simple pattern recognition, these benchmarks test performance on formal proofs, competition-level Olympiad problems, multi-step algebra, and discrete mathematics, where accuracy and logical consistency are critical.

By aggregating standardized scores from various sources this leaderboard highlights the models that perform best in mathematical problem-solving, multi-step reasoning, and numerical accuracy.

Looking for a balanced model across all domains? Mathematical performance contributes 20% to the composite score. Visit the Overall LLM Ranking to see how top math models compare in coding, agentic capabilities, and broader logical reasoning.

Last updated: November 5, 2026
Rank Model Name Relative Quality Score
1 Gemini 4 Carbon
Quality of Math 98.91%

98.91
2 Gemini 4 Argon
Quality of Math 97.26%

97.26
3 GPT-6.1 Terra
Quality of Math 97.01%

97.01
4 Claude Fable 5.1
Quality of Math 97.00%

97.00
5 Claude Opus 5
Quality of Math 94.96%

94.96
6 Claude Opus 5.5
Quality of Math 94.90%

94.90
7 Claude Fable 5
Quality of Math 94.32%

94.32
8 Muse Spark 1.2
Quality of Math 93.76%

93.76
9 MiMo-V2.6-Pro
Quality of Math 93.45%

93.45
10 Muse Spark 1.1
Quality of Math 92.18%

92.18
11 Claude Sonnet 5.5
Quality of Math 91.51%

91.51
12 Muse Wonder
Quality of Math 91.40%

91.40
13 Muse Spark 1.3
Quality of Math 90.64%

90.64
14 GPT-6.1 Sol
Quality of Math 90.55%

90.55
15 GPT-5.5
Quality of Math 90.51%

90.51
16 Gemini 3.8 Flash
Quality of Math 90.17%

90.17
17 GPT-6 Sol
Quality of Math 89.88%

89.88
18 Kimi K3
Quality of Math 89.85%

89.85
19 Claude Opus 4.8
Quality of Math 89.82%

89.82
20 Step 5
Quality of Math 89.70%

89.70
21 MiMo-V2.6-Flash
Quality of Math 89.24%

89.24
22 GLM-5.3
Quality of Math 89.12%

89.12
23 Gemini 3.7 Flash
Quality of Math 89.12%

89.12
24 Muse Spark
Quality of Math 89.06%

89.06
25 Claude Opus 4.6
Quality of Math 88.64%

88.64
26 Claude Opus 4.7
Quality of Math 88.17%

88.17
27 GPT-5.6 Sol
Quality of Math 88.13%

88.13
28 GPT-6 Astra
Quality of Math 88.05%

88.05
29 Qwen3.7
Quality of Math 87.66%

87.66
30 Gemini 3 Pro
Quality of Math 87.65%

87.65
31 GPT-5.5 Pro
Quality of Math 87.64%

87.64
32 MiMo-V2-Pro
Quality of Math 87.36%

87.36
33 GPT-5.4
Quality of Math 87.21%

87.21
34 Gemini 3.1 Pro
Quality of Math 86.85%

86.85
35 Kimi K2.5
Quality of Math 86.63%

86.63
36 Hy3
Quality of Math 86.40%

86.40
37 GLM-5
Quality of Math 86.27%

86.27
38 Muse Glimmer
Quality of Math 86.27%

86.27
39 GLM-5.2
Quality of Math 86.01%

86.01
40 Grok-4.1
Quality of Math 85.96%

85.96
41 Kimi K3.1 Pro
Quality of Math 85.92%

85.92
42 GPT-6 Luna
Quality of Math 85.89%

85.89
43 Hy4
Quality of Math 85.60%

85.60
44 DeepSeek-V4.1-Flash
Quality of Math 85.38%

85.38
45 MiMo-V2.5
Quality of Math 85.34%

85.34
46 GLM-5V-Turbo
Quality of Math 85.34%

85.34
47 GPT-5.6 Terra
Quality of Math 85.10%

85.10
48 Qwen3.8
Quality of Math 85.00%

85.00
49 DeepSeek-V4-Pro
Quality of Math 84.98%

84.98
50 MiMo-V2-Omni
Quality of Math 84.71%

84.71
51 Gemini 3.6 Flash
Quality of Math 84.60%

84.60
52 Claude Sonnet 4.6
Quality of Math 84.45%

84.45
53 Claude Sonnet 5
Quality of Math 84.31%

84.31
54 GLM-5.3-Flash
Quality of Math 83.72%

83.72
55 GPT-5.2 Pro
Quality of Math 83.71%

83.71
56 GPT-5.2
Quality of Math 83.63%

83.63
57 Kimi K2
Quality of Math 83.28%

83.28
58 MiniMax M2.7
Quality of Math 82.84%

82.84
59 Qwen3.5-122B-A10B
Quality of Math 82.65%

82.65
60 GPT-5.6 Luna
Quality of Math 82.07%

82.07
61 Kimi K2.6
Quality of Math 82.02%

82.02
62 GLM-4.7
Quality of Math 81.72%

81.72
63 GPT-5.1
Quality of Math 81.71%

81.71
64 Qwen3.5-27B
Quality of Math 81.64%

81.64
65 DeepSeek-V4-Flash
Quality of Math 81.40%

81.40
66 Seed 2.0 Pro
Quality of Math 81.32%

81.32
67 Gemini 3 Flash
Quality of Math 80.92%

80.92
68 Grok 4.6
Quality of Math 80.89%

80.89
69 Grok 4.5
Quality of Math 80.78%

80.78
70 Gemini 3.5 Flash
Quality of Math 80.77%

80.77
71 Gemma 4 31B
Quality of Math 80.77%

80.77
72 Beam 1.0
Quality of Math 80.44%

80.44
73 Qwen3.8-27B
Quality of Math 80.38%

80.38
74 ChatGPT-4o
Quality of Math 79.88%

79.88
75 Step-3.5-Flash
Quality of Math 79.61%

79.61
76 ERNIE 5.0
Quality of Math 79.52%

79.52
77 DeepSeek-V3.2
Quality of Math 79.29%

79.29
78 Claude Haiku 5.5
Quality of Math 79.08%

79.08
79 Grok-4
Quality of Math 79.06%

79.06
80 DeepSeek V3.1 Terminus
Quality of Math 78.94%

78.94
81 LongCat-Flash
Quality of Math 78.90%

78.90
82 Mistral Medium 3.1
Quality of Math 78.63%

78.63
83 Solar Pro 4
Quality of Math 78.59%

78.59
84 MiMo-V2.5-Pro
Quality of Math 78.49%

78.49
85 MiniMax M2.5
Quality of Math 78.32%

78.32
86 Qwen3.7-Plus
Quality of Math 78.22%

78.22
87 Qwen3 235B A22B
Quality of Math 78.20%

78.20
88 Qwen3.5-35B-A3B
Quality of Math 77.89%

77.89
89 Qwen3.6 Plus
Quality of Math 77.78%

77.78
90 Gemma 4 26B-A4B
Quality of Math 77.56%

77.56
91 Trinity Large
Quality of Math 76.76%

76.76
92 Qwen3.8-Flash-Next
Quality of Math 76.37%

76.37
93 Muse Spark Open
Quality of Math 76.27%

76.27
94 Grok 4.7
Quality of Math 76.26%

76.26
95 Qwen3 VL 235B A22B
Quality of Math 76.24%

76.24
96 Grok-4.20
Quality of Math 75.67%

75.67
97 Qwen3-Coder 480B A35B
Quality of Math 75.66%

75.66
98 GPT-5
Quality of Math 75.63%

75.63
99 Grok 4 Fast
Quality of Math 75.53%

75.53
100 Inkling-Small
Quality of Math 75.44%

75.44