The Coding Leaderboard evaluates LLMs on their software engineering capabilities. From generating production-ready code across major programming languages to designing algorithms, debugging complex functions, and translating scientific concepts into executable scripts, this category measures a model’s ability to handle a broad range of coding and software development tasks.

Scores in this category are compiled from leading coding benchmarks highlighting models that perform well on code generation, debugging, problem-solving, and technical accuracy.

Need more than code generation? Coding performance accounts for 30% of our composite rating. Explore the Overall LLM Ranking to see which top coding models also excel at autonomous agent workflows, advanced mathematics, and complex reasoning.

Last updated: October 9, 2026
Rank Model Name Relative Quality Score
1 Claude Opus 5.5
Quality of Coding 99.73%

99.73
2 Claude Sonnet 5.5
Quality of Coding 95.51%

95.51
3 GPT-6 Astra
Quality of Coding 92.68%

92.68
4 GPT-5.6 Sol
Quality of Coding 90.88%

90.88
5 Gemini 4 Argon
Quality of Coding 89.42%

89.42
6 ERNIE 5.0
Quality of Coding 89.40%

89.40
7 Claude Fable 5.1
Quality of Coding 89.36%

89.36
8 GPT-6.1 Sol
Quality of Coding 88.91%

88.91
9 Claude Fable 5
Quality of Coding 88.43%

88.43
10 Kimi K3
Quality of Coding 86.57%

86.57
11 Muse Spark 1.3
Quality of Coding 86.51%

86.51
12 Claude Opus 5
Quality of Coding 86.45%

86.45
13 ChatGPT-4o
Quality of Coding 86.01%

86.01
14 GPT-5.6 Terra
Quality of Coding 85.68%

85.68
15 Qwen3 VL 235B A22B
Quality of Coding 85.42%

85.42
16 GPT-5.5
Quality of Coding 85.28%

85.28
17 GPT-5.5 Instant
Quality of Coding 84.73%

84.73
18 GPT-6 Sol
Quality of Coding 84.51%

84.51
19 MiMo-V2.6-Pro
Quality of Coding 84.39%

84.39
20 GLM-5.3
Quality of Coding 84.28%

84.28
21 Mistral
Quality of Coding 84.09%

84.09
22 GPT-5.6 Luna
Quality of Coding 83.66%

83.66
23 DeepSeek-V4.1-Flash
Quality of Coding 83.36%

83.36
24 Qwen3.8
Quality of Coding 82.35%

82.35
25 Grok 4.7
Quality of Coding 82.17%

82.17
26 Claude Opus 4.8
Quality of Coding 81.28%

81.28
27 Claude Opus 4.7
Quality of Coding 80.48%

80.48
28 Gemini 3.7 Flash
Quality of Coding 80.41%

80.41
29 Ling 3.1 Flash
Quality of Coding 80.23%

80.23
30 Grok 4.6
Quality of Coding 80.01%

80.01
31 DeepSeek-V4-Pro
Quality of Coding 79.52%

79.52
32 Gemini 3.8 Flash
Quality of Coding 79.47%

79.47
33 Claude Haiku 5.5
Quality of Coding 79.20%

79.20
34 GLM-5.3-Flash
Quality of Coding 79.14%

79.14
35 GLM-5.2
Quality of Coding 79.06%

79.06
36 Step 5
Quality of Coding 78.64%

78.64
37 Muse Spark 1.1
Quality of Coding 78.59%

78.59
38 MiMo-V2.6-Flash
Quality of Coding 78.53%

78.53
39 INTELLECT-3
Quality of Coding 78.50%

78.50
40 GLM-4.6V
Quality of Coding 78.50%

78.50
41 Claude Sonnet 5
Quality of Coding 78.46%

78.46
42 Mistral Small
Quality of Coding 77.91%

77.91
43 Hy4
Quality of Coding 77.64%

77.64
44 Ling-flash-2.0
Quality of Coding 77.61%

77.61
45 Mistral Large 4
Quality of Coding 77.47%

77.47
46 GPT-6 Luna
Quality of Coding 77.46%

77.46
47 Claude Opus 4.6
Quality of Coding 77.44%

77.44
48 GLM-4.5V
Quality of Coding 76.44%

76.44
49 Qwen2.5
Quality of Coding 76.44%

76.44
50 DeepSeek-V4-Flash
Quality of Coding 76.39%

76.39
51 Qwen3.8-Flash-Next
Quality of Coding 76.12%

76.12
52 Muse Spark 1.2
Quality of Coding 75.86%

75.86
53 Kimi K2.6
Quality of Coding 75.62%

75.62
54 Grok 4.5
Quality of Coding 75.52%

75.52
55 Qwen3.7
Quality of Coding 75.31%

75.31
56 Qwen3.8-27B
Quality of Coding 75.20%

75.20
57 Gemini 3.5 Flash
Quality of Coding 74.77%

74.77
58 Ring-flash-2.0
Quality of Coding 74.67%

74.67
59 Gemini 3.6 Flash
Quality of Coding 74.45%

74.45
60 GPT-5 nano
Quality of Coding 73.93%

73.93
61 Gemini 3.1 Pro
Quality of Coding 73.82%

73.82
62 Claude Opus 4.5
Quality of Coding 73.47%

73.47
63 Olmo 3.1 32B
Quality of Coding 73.34%

73.34
64 Qwen3.7-Plus
Quality of Coding 73.32%

73.32
65 Grok Build 0.1
Quality of Coding 73.24%

73.24
66 Claude Sonnet 4.6
Quality of Coding 72.96%

72.96
67 Muse Spark
Quality of Coding 71.17%

71.17
68 MiMo-V2.5-Pro
Quality of Coding 70.76%

70.76
69 Hy3
Quality of Coding 70.71%

70.71
70 Qwen3.6 Plus
Quality of Coding 70.67%

70.67
71 Olmo 3 32B Think
Quality of Coding 70.25%

70.25
72 MiniMax M3
Quality of Coding 69.86%

69.86
73 GPT-5.4
Quality of Coding 69.82%

69.82
74 Seed 2.0 Pro
Quality of Coding 69.49%

69.49
75 Mistral Large
Quality of Coding 69.37%

69.37
76 Qwen3.6
Quality of Coding 69.11%

69.11
77 GLM-5.1
Quality of Coding 68.51%

68.51
78 DeepSeek V3.1 Terminus
Quality of Coding 68.32%

68.32
79 Olmo 3.1 32B Think
Quality of Coding 67.45%

67.45
80 Qwen3.6-27B
Quality of Coding 67.12%

67.12
81 GPT-5 mini
Quality of Coding 67.09%

67.09
82 GPT-5.2
Quality of Coding 67.02%

67.02
83 Kimi K2.7 Code
Quality of Coding 67.02%

67.02
84 GPT-5.3 Codex
Quality of Coding 67.02%

67.02
85 GPT-5.4 nano
Quality of Coding 66.64%

66.64
86 Inkling
Quality of Coding 66.44%

66.44
87 MiMo-V2-Omni
Quality of Coding 66.33%

66.33
88 LongCat-Flash
Quality of Coding 65.81%

65.81
89 Kimi K2
Quality of Coding 65.61%

65.61
90 Seed 2.1 Pro
Quality of Coding 65.55%

65.55
91 Llama 3.1 Nemotron 70B
Quality of Coding 65.54%

65.54
92 GPT-5.4 mini
Quality of Coding 65.40%

65.40
93 Grok 4.3
Quality of Coding 65.07%

65.07
94 Step-3.5-Flash
Quality of Coding 64.37%

64.37
95 Inkling-Small
Quality of Coding 64.36%

64.36
96 MiMo-V2.5
Quality of Coding 64.31%

64.31
97 Claude Opus 4
Quality of Coding 64.21%

64.21
98 Granite 4.2 30B
Quality of Coding 63.90%

63.90
99 MiniMax M2.7
Quality of Coding 63.47%

63.47
100 Gemini 3 Pro
Quality of Coding 63.27%

63.27