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: October 9, 2026
Rank Model Name Relative Quality Score
1 Gemini 4 Argon
Quality of Math 100.00%

100.00
2 Gemini 3.8 Flash
Quality of Math 97.48%

97.48
3 Claude Fable 5.1
Quality of Math 97.45%

97.45
4 Muse Spark 1.3
Quality of Math 96.87%

96.87
5 Claude Opus 5
Quality of Math 96.49%

96.49
6 Gemini 3.7 Flash
Quality of Math 96.15%

96.15
7 Claude Fable 5
Quality of Math 95.98%

95.98
8 Claude Opus 5.5
Quality of Math 94.89%

94.89
9 Gemini 3.6 Flash
Quality of Math 93.61%

93.61
10 Claude Opus 4.6
Quality of Math 93.41%

93.41
11 Gemini 3.1 Pro
Quality of Math 93.14%

93.14
12 Muse Spark 1.2
Quality of Math 93.06%

93.06
13 Claude Sonnet 5.5
Quality of Math 92.97%

92.97
14 MiMo-V2.6-Pro
Quality of Math 92.94%

92.94
15 Claude Opus 4.7
Quality of Math 92.57%

92.57
16 Claude Opus 4.8
Quality of Math 92.42%

92.42
17 Qwen3.7
Quality of Math 92.32%

92.32
18 GPT-5.5
Quality of Math 92.31%

92.31
19 GLM-5.2
Quality of Math 92.26%

92.26
20 Step 5
Quality of Math 92.15%

92.15
21 GLM-5.3
Quality of Math 92.12%

92.12
22 Grok 4.5
Quality of Math 92.08%

92.08
23 Qwen3.6
Quality of Math 91.84%

91.84
24 Muse Spark 1.1
Quality of Math 91.76%

91.76
25 Kimi K3
Quality of Math 91.65%

91.65
26 DeepSeek-V4-Pro
Quality of Math 91.33%

91.33
27 GPT-6 Sol
Quality of Math 91.18%

91.18
28 Grok 4.6
Quality of Math 90.40%

90.40
29 GPT-5.4
Quality of Math 90.30%

90.30
30 Grok-4.20
Quality of Math 90.11%

90.11
31 Grok 4.7
Quality of Math 90.10%

90.10
32 GPT-6.1 Sol
Quality of Math 89.78%

89.78
33 MiMo-V2.6-Flash
Quality of Math 89.64%

89.64
34 Claude Sonnet 5
Quality of Math 89.55%

89.55
35 GPT-5.6 Sol
Quality of Math 89.53%

89.53
36 GPT-6 Luna
Quality of Math 89.40%

89.40
37 Qwen3.7-Plus
Quality of Math 89.37%

89.37
38 DeepSeek-V4.1-Flash
Quality of Math 89.31%

89.31
39 Muse Spark
Quality of Math 89.13%

89.13
40 Mistral Large 4
Quality of Math 89.12%

89.12
41 Seed 2.1 Pro
Quality of Math 88.94%

88.94
42 GPT-5.2
Quality of Math 88.92%

88.92
43 GPT-6 Astra
Quality of Math 88.77%

88.77
44 GLM-5.3-Flash
Quality of Math 88.46%

88.46
45 Gemini 3 Pro
Quality of Math 88.18%

88.18
46 Qwen3.8
Quality of Math 88.14%

88.14
47 Gemini 3 Flash
Quality of Math 87.86%

87.86
48 Kimi K2.6
Quality of Math 87.84%

87.84
49 MiMo-V2-Pro
Quality of Math 87.76%

87.76
50 DeepSeek-V4-Flash
Quality of Math 87.63%

87.63
51 Qwen3.6 Plus
Quality of Math 87.60%

87.60
52 Kimi K2.5
Quality of Math 87.31%

87.31
53 Muse Glimmer
Quality of Math 87.28%

87.28
54 Hy3
Quality of Math 86.99%

86.99
55 GPT-5.6 Terra
Quality of Math 86.93%

86.93
56 Hy4
Quality of Math 86.89%

86.89
57 GLM-5
Quality of Math 86.66%

86.66
58 Grok-4.1
Quality of Math 86.50%

86.50
59 Claude Sonnet 4.6
Quality of Math 85.88%

85.88
60 MiMo-V2.5
Quality of Math 85.87%

85.87
61 GLM-5V-Turbo
Quality of Math 85.87%

85.87
62 Gemini 3.5 Flash
Quality of Math 85.86%

85.86
63 Qwen3.5-397B-A17B
Quality of Math 85.82%

85.82
64 Claude Opus 4.5
Quality of Math 85.81%

85.81
65 MiMo-V2-Omni
Quality of Math 85.09%

85.09
66 GLM-5.1
Quality of Math 84.86%

84.86
67 Grok 4.3
Quality of Math 84.20%

84.20
68 Kimi K2
Quality of Math 84.02%

84.02
69 Inkling
Quality of Math 83.63%

83.63
70 Inkling-Small
Quality of Math 83.61%

83.61
71 Qwen3.5-122B-A10B
Quality of Math 83.30%

83.30
72 MiniMax M2.7
Quality of Math 83.20%

83.20
73 Grok 4.1 Fast
Quality of Math 83.20%

83.20
74 GPT-5.6 Luna
Quality of Math 83.00%

83.00
75 Gemini 3.5 Flash-Lite
Quality of Math 82.97%

82.97
76 Qwen3.8-27B
Quality of Math 82.87%

82.87
77 Qwen3.5-27B
Quality of Math 82.36%

82.36
78 GPT-5.1
Quality of Math 82.26%

82.26
79 GLM-4.7
Quality of Math 82.12%

82.12
80 Seed 2.0 Pro
Quality of Math 82.11%

82.11
81 Gemma 4 31B
Quality of Math 81.55%

81.55
82 Qwen3.8-Flash-Next
Quality of Math 81.06%

81.06
83 Grok Build 0.1
Quality of Math 80.74%

80.74
84 Qwen3.6-27B
Quality of Math 80.61%

80.61
85 ChatGPT-4o
Quality of Math 80.53%

80.53
86 Qwen3
Quality of Math 80.46%

80.46
87 GPT-5
Quality of Math 80.46%

80.46
88 Step-3.5-Flash
Quality of Math 80.23%

80.23
89 ERNIE 5.0
Quality of Math 80.06%

80.06
90 DeepSeek-V3.2
Quality of Math 79.91%

79.91
91 DeepSeek-V3.2-Speciale
Quality of Math 79.78%

79.78
92 LongCat-Flash
Quality of Math 79.67%

79.67
93 Grok-4
Quality of Math 79.59%

79.59
94 DeepSeek V3.1 Terminus
Quality of Math 79.43%

79.43
95 MiMo-V2.5-Pro
Quality of Math 79.17%

79.17
96 Mistral
Quality of Math 79.12%

79.12
97 Qwen3 235B A22B
Quality of Math 78.89%

78.89
98 MiniMax M2.5
Quality of Math 78.81%

78.81
99 Qwen3.5-35B-A3B
Quality of Math 78.50%

78.50
100 Gemma 4 26B-A4B
Quality of Math 78.23%

78.23