LLM Rankings: Top AI Models – Lewoniewski.info

This leaderboard presents the top 100 Large Language Models (LLMs) ranked by their Relative Quality Score. Unlike single-source lists, this ranking utilizes a unified, synthetic aggregation model to deliver a comprehensive and unbiased picture of current AI capabilities.

The ranking includes only those AI models that have been evaluated by at least two widely recognized independent benchmarking sources. Each model’s overall score is calculated as a composite across four key domains:

💻 Coding (40%) — measuring programming ability, syntax accuracy, and algorithmic logic;
🧮 Math (15%) — testing quantitative problem-solving;
🔬 Reasoning (20%) — assessing logical thinking and multi-step deduction;
✍️ Instruction Following (25%) — evaluating how well the model follows custom constraints and user instructions.

Clicking on any model expands details across all four sub-aspects.

Rank	Model Name	Relative Quality	Score
1	Claude Opus 5	Overall Quality 94.59%	94.59
Coding 94.07% Math 97.20% Reasoning 94.75% Instr. Following 97.36%
2	Claude Fable 5	Overall Quality 94.36%	94.36
Coding 98.03% Math 96.43% Reasoning 93.18% Instr. Following 91.32%
3	Claude Fable 5.1	Overall Quality 94.16%	94.16
Coding 91.75% Math 94.59% Reasoning 96.63% Instr. Following 95.87%
4	GPT-6 Astra	Overall Quality 93.99%	93.99
Coding 95.24% Math 81.53% Reasoning 97.89% Instr. Following —
5	Kimi K3	Overall Quality 93.91%	93.91
Coding 96.05% Math 93.83% Reasoning 89.59% Instr. Following 95.70%
6	GPT-5.6 Sol	Overall Quality 92.81%	92.81
Coding 95.93% Math 88.35% Reasoning 91.65% Instr. Following 92.42%
7	GLM-5.3	Overall Quality 92.65%	92.65
Coding 94.50% Math 94.83% Reasoning 86.43% Instr. Following 94.38%
8	Gemini 3.8 Flash	Overall Quality 92.41%	92.41
Coding 93.59% Math 95.62% Reasoning 87.30% Instr. Following 96.20%
9	Gemini 3.7 Flash	Overall Quality 92.28%	92.28
Coding 91.42% Math 99.22% Reasoning 87.41% Instr. Following 95.21%
10	Muse Spark 1.3	Overall Quality 91.66%	91.66
Coding 89.94% Math — Reasoning 90.76% Instr. Following —
11	GPT-5.5	Overall Quality 91.16%	91.16
Coding 92.16% Math 91.80% Reasoning 86.34% Instr. Following 92.88%
12	Claude Opus 4.8	Overall Quality 90.30%	90.30
Coding 94.22% Math 92.00% Reasoning 88.84% Instr. Following 88.58%
13	Muse Spark 1.1	Overall Quality 90.04%	90.04
Coding 91.76% Math 92.20% Reasoning 86.39% Instr. Following 92.73%
14	Qwen3.7 Max	Overall Quality 89.80%	89.80
Coding 90.10% Math 93.85% Reasoning 82.18% Instr. Following 93.49%
15	Gemini 3.1 Pro	Overall Quality 89.74%	89.74
Coding 88.72% Math 93.19% Reasoning 85.42% Instr. Following 93.65%
16	Qwen3.8 Max	Overall Quality 89.26%	89.26
Coding 91.14% Math 88.08% Reasoning 86.14% Instr. Following 93.88%
17	GPT-5.6 Terra	Overall Quality 89.06%	89.06
Coding 92.78% Math 86.29% Reasoning 85.36% Instr. Following 89.03%
18	Grok 4.6	Overall Quality 88.44%	88.44
Coding 90.30% Math 83.88% Reasoning 86.53% Instr. Following 90.74%
19	Grok 4.5	Overall Quality 88.37%	88.37
Coding 90.28% Math 92.18% Reasoning 82.24% Instr. Following 92.07%
20	Claude Opus 4.7	Overall Quality 88.32%	88.32
Coding 93.68% Math 92.46% Reasoning 82.13% Instr. Following 88.59%
21	GPT-5.4	Overall Quality 88.31%	88.31
Coding 89.02% Math 89.83% Reasoning 81.86% Instr. Following 92.30%
22	Claude Sonnet 5	Overall Quality 88.02%	88.02
Coding 90.48% Math 89.42% Reasoning 82.72% Instr. Following 91.90%
23	GLM-5.3-Flash	Overall Quality 87.29%	87.29
Coding 90.44% Math 88.24% Reasoning 83.14% Instr. Following 91.90%
24	DeepSeek-V4-Pro-0813	Overall Quality 87.20%	87.20
Coding 87.47% Math 90.67% Reasoning 84.70% Instr. Following 89.98%
25	Qwen3.6 Max Preview	Overall Quality 86.89%	86.89
Coding 89.73% Math 91.55% Reasoning 71.37% Instr. Following 90.06%
26	Muse Spark 1.2	Overall Quality 86.66%	86.66
Coding 89.83% Math — Reasoning 76.30% Instr. Following 94.21%
27	Gemini 3.6 Flash	Overall Quality 86.55%	86.55
Coding 86.99% Math 96.09% Reasoning 79.19% Instr. Following 93.06%
28	Gemini 3.5 Flash	Overall Quality 86.51%	86.51
Coding 85.84% Math 87.21% Reasoning 81.89% Instr. Following 91.60%
29	Claude Opus 4.6	Overall Quality 86.40%	86.40
Coding 91.11% Math 94.48% Reasoning 80.62% Instr. Following 87.32%
30	GLM-5.2	Overall Quality 85.95%	85.95
Coding 86.91% Math 91.46% Reasoning 80.46% Instr. Following 90.11%
31	Kimi K2.6	Overall Quality 85.94%	85.94
Coding 86.91% Math 88.71% Reasoning 78.50% Instr. Following 90.10%
32	GPT-5.6 Luna	Overall Quality 85.33%	85.33
Coding 87.90% Math 83.62% Reasoning 79.77% Instr. Following 88.26%
33	Qwen3.7-Plus	Overall Quality 85.16%	85.16
Coding 84.96% Math 89.47% Reasoning 77.12% Instr. Following 90.02%
34	GPT-5.2 Codex	Overall Quality 84.66%	84.66
Coding 84.36% Math 88.05% Reasoning 77.29% Instr. Following 91.41%
35	Muse Spark	Overall Quality 84.40%	84.40
Coding 84.54% Math 88.71% Reasoning 77.89% Instr. Following 91.14%
36	GLM-5.1	Overall Quality 83.60%	83.60
Coding 85.68% Math 86.52% Reasoning 70.69% Instr. Following 90.79%
37	Gemini 3 Pro	Overall Quality 83.59%	83.59
Coding 83.49% Math 88.81% Reasoning 77.02% Instr. Following 89.76%
38	DeepSeek-V4-Flash-0731	Overall Quality 83.22%	83.22
Coding 81.83% Math 87.47% Reasoning 78.72% Instr. Following 89.05%
39	Gemini 3 Flash	Overall Quality 83.01%	83.01
Coding 81.12% Math 88.50% Reasoning 74.65% Instr. Following 91.52%
40	Grok 4.3	Overall Quality 82.91%	82.91
Coding 85.74% Math 82.94% Reasoning 66.36% Instr. Following 88.66%
41	Inkling	Overall Quality 82.78%	82.78
Coding 87.52% Math 91.86% Reasoning 71.08% Instr. Following 84.46%
42	MiMo-V2.5-Pro	Overall Quality 82.68%	82.68
Coding 85.27% Math 81.55% Reasoning 65.67% Instr. Following 93.76%
43	MiniMax M3	Overall Quality 82.54%	82.54
Coding 82.72% Math 77.76% Reasoning 78.80% Instr. Following 90.86%
44	Qwen3.6 Plus	Overall Quality 82.05%	82.05
Coding 82.89% Math 88.02% Reasoning 70.77% Instr. Following 87.71%
45	Qwen3.8-27B	Overall Quality 81.65%	81.65
Coding 84.22% Math 83.39% Reasoning 77.36% Instr. Following 86.12%
46	Hy3	Overall Quality 81.60%	81.60
Coding 83.70% Math 87.73% Reasoning 75.49% Instr. Following 83.57%
47	Claude Sonnet 4.6	Overall Quality 81.58%	81.58
Coding 87.11% Math 85.71% Reasoning 72.30% Instr. Following 84.62%
48	Claude Opus 4.5	Overall Quality 81.58%	81.58
Coding 87.10% Math 85.03% Reasoning 71.31% Instr. Following 85.85%
49	Seed 2.0 Pro	Overall Quality 81.50%	81.50
Coding 84.43% Math 83.19% Reasoning 72.46% Instr. Following 86.94%
50	Claude Sonnet 4.5	Overall Quality 81.24%	81.24
Coding 85.17% Math 79.04% Reasoning 62.17% Instr. Following 91.74%
51	Qwen3.5-397B-A17B	Overall Quality 81.08%	81.08
Coding 80.16% Math 85.83% Reasoning 70.70% Instr. Following 89.05%
52	Qwen3.8-Flash-Next	Overall Quality 80.66%	80.66
Coding 77.67% Math 77.23% Reasoning 82.10% Instr. Following —
53	Kimi K2.5	Overall Quality 80.48%	80.48
Coding 81.09% Math 87.93% Reasoning 71.57% Instr. Following 85.83%
54	Inkling-Small	Overall Quality 80.48%	80.48
Coding 83.48% Math 84.65% Reasoning 72.64% Instr. Following 80.00%
55	MiMo-V2-Pro	Overall Quality 80.28%	80.28
Coding 82.45% Math 87.95% Reasoning 69.91% Instr. Following 85.99%
56	Gemma 4 31B	Overall Quality 80.10%	80.10
Coding 83.78% Math 83.44% Reasoning 64.29% Instr. Following 89.87%
57	Claude Opus 4.1	Overall Quality 79.98%	79.98
Coding 84.31% Math 76.41% Reasoning 60.22% Instr. Following 90.91%
58	GLM-5	Overall Quality 79.94%	79.94
Coding 82.36% Math 86.54% Reasoning 67.42% Instr. Following 87.48%
59	GPT-5.1 Thinking	Overall Quality 79.79%	79.79
Coding 80.34% Math 83.21% Reasoning 68.19% Instr. Following 88.37%
60	Muse Glimmer	Overall Quality 79.23%	79.23
Coding 85.78% Math 88.11% Reasoning 60.57% Instr. Following 83.80%
61	MiniMax M2.7	Overall Quality 79.17%	79.17
Coding 80.55% Math 83.26% Reasoning 68.30% Instr. Following 84.66%
62	MiMo-V2.5	Overall Quality 78.62%	78.62
Coding 81.83% Math 86.23% Reasoning 66.41% Instr. Following 83.75%
63	GPT-5.3 Codex	Overall Quality 78.54%	78.54
Coding 72.68% Math 58.76% Reasoning 77.75% Instr. Following 88.73%
64	GPT-5	Overall Quality 77.87%	77.87
Coding 78.91% Math 81.05% Reasoning 66.38% Instr. Following 84.44%
65	Gemma 4 26B-A4B	Overall Quality 77.09%	77.09
Coding 82.03% Math 80.32% Reasoning 56.72% Instr. Following 86.67%
66	Claude Opus 4	Overall Quality 76.58%	76.58
Coding 81.10% Math 72.82% Reasoning 57.02% Instr. Following 88.43%
67	Qwen3.5-122B-A10B	Overall Quality 76.29%	76.29
Coding 74.37% Math 83.26% Reasoning 65.77% Instr. Following 84.66%
68	LongCat-Flash-Thinking-2601	Overall Quality 75.91%	75.91
Coding 78.87% Math 79.04% Reasoning 68.29% Instr. Following 79.17%
69	GLM-4.7	Overall Quality 75.90%	75.90
Coding 76.18% Math 82.75% Reasoning 66.21% Instr. Following 83.60%
70	Gemini 3.1 Flash-Lite	Overall Quality 75.73%	75.73
Coding 78.74% Math 75.66% Reasoning 57.82% Instr. Following 85.08%
71	MiMo-V2-Omni	Overall Quality 75.34%	75.34
Coding 77.94% Math 85.45% Reasoning 59.52% Instr. Following 83.17%
72	Qwen3.5-27B	Overall Quality 75.24%	75.24
Coding 72.56% Math 83.03% Reasoning 64.74% Instr. Following 83.85%
73	DeepSeek-V3.2	Overall Quality 74.89%	74.89
Coding 77.86% Math 80.93% Reasoning 63.90% Instr. Following 79.83%
74	MiniMax M2.5	Overall Quality 74.77%	74.77
Coding 75.96% Math 79.19% Reasoning 64.09% Instr. Following 79.96%
75	Kimi K2-Thinking-0905	Overall Quality 74.64%	74.64
Coding 75.57% Math 83.63% Reasoning 65.04% Instr. Following 79.08%
76	GPT-5 mini	Overall Quality 74.32%	74.32
Coding 77.35% Math 74.85% Reasoning 57.77% Instr. Following 80.19%
77	DeepSeek V3.1 Terminus	Overall Quality 73.69%	73.69
Coding 80.98% Math 79.19% Reasoning 52.40% Instr. Following 78.25%
78	Step-3.5-Flash	Overall Quality 73.32%	73.32
Coding 73.81% Math 80.04% Reasoning 65.03% Instr. Following 78.31%
79	Gemini 3.5 Flash-Lite	Overall Quality 73.32%	73.32
Coding 78.02% Math 86.23% Reasoning 56.63% Instr. Following 88.10%
80	o3	Overall Quality 73.21%	73.21
Coding 73.61% Math 78.89% Reasoning 59.42% Instr. Following 81.97%
81	Claude Sonnet 4	Overall Quality 72.82%	72.82
Coding 77.96% Math 68.55% Reasoning 52.29% Instr. Following 83.64%
82	Qwen3 Max	Overall Quality 72.63%	72.63
Coding 72.24% Math 75.54% Reasoning 59.72% Instr. Following 84.75%
83	GLM-5V-Turbo	Overall Quality 72.19%	72.19
Coding 79.15% Math 86.54% Reasoning 44.51% Instr. Following 80.18%
84	MiniMax M2.1	Overall Quality 72.16%	72.16
Coding 72.60% Math 74.53% Reasoning 61.89% Instr. Following 79.47%
85	MiMo-V2-Flash	Overall Quality 72.11%	72.11
Coding 72.05% Math 74.80% Reasoning 61.81% Instr. Following 80.20%
86	Grok-4	Overall Quality 71.89%	71.89
Coding 72.06% Math 80.46% Reasoning 66.13% Instr. Following 74.21%
87	Kimi K2.7 Code	Overall Quality 71.56%	71.56
Coding 68.86% Math 55.10% Reasoning 71.81% Instr. Following 71.83%
88	Qwen3.5-35B-A3B	Overall Quality 71.26%	71.26
Coding 68.12% Math 78.84% Reasoning 61.36% Instr. Following 81.20%
89	ERNIE 5.0	Overall Quality 70.95%	70.95
Coding 83.05% Math 81.17% Reasoning 55.27% Instr. Following 72.07%
90	Qwen3-235B-A22B-Instruct-2507	Overall Quality 70.74%	70.74
Coding 72.66% Math 75.06% Reasoning 51.32% Instr. Following 83.80%
91	GPT-4.5	Overall Quality 70.49%	70.49
Coding 71.91% Math 71.95% Reasoning 46.73% Instr. Following 87.27%
92	Mistral Medium 3.5	Overall Quality 70.43%	70.43
Coding 75.36% Math 73.25% Reasoning 51.63% Instr. Following 83.44%
93	Qwen3-Coder 480B A35B Instruct	Overall Quality 70.21%	70.21
Coding 74.39% Math 75.74% Reasoning 47.98% Instr. Following 78.51%
94	Grok-3 Mini	Overall Quality 70.16%	70.16
Coding 69.78% Math 75.04% Reasoning 61.75% Instr. Following 74.71%
95	DeepSeek-V3.2-Exp	Overall Quality 69.87%	69.87
Coding 75.08% Math 77.21% Reasoning 54.40% Instr. Following 76.41%
96	Gemini 2.5 Pro	Overall Quality 69.41%	69.41
Coding 71.62% Math 76.35% Reasoning 59.34% Instr. Following 76.84%
97	Qwen3.6-27B	Overall Quality 69.24%	69.24
Coding 58.97% Math 79.94% Reasoning 64.40% Instr. Following 78.87%
98	Trinity Large Thinking	Overall Quality 69.23%	69.23
Coding 76.60% Math 77.46% Reasoning 50.07% Instr. Following 72.05%
99	Grok 4 Fast	Overall Quality 68.78%	68.78
Coding 71.48% Math 77.81% Reasoning 59.76% Instr. Following 72.59%
100	Solar Pro 4	Overall Quality 67.91%	67.91
Coding 57.55% Math 76.59% Reasoning 68.81% Instr. Following —