Hallucination

28 scenarios across 3 subcategories. 155 models scored.

Subcategories

Subcategory Avg Score Best Model Best Score
False Positives 57.58% Nemotron 3 Super 99.06%
Content Invention 96.74% Qwen3.7 Max 100.00%
Output Corruption 99.13% Claude Opus 4.6 (Reasoning) 100.00%

Model Leaderboard

All models ranked by their Hallucination category score.

# Model Hallucination False Positives Content Invention Output Corruption Overall
1 Stealth: Aurora Alpha 99.93% 99.93% 83.79%
2 Nemotron 3 Super 99.69% 99.06% 100.00% 100.00% 84.56%
3 o4 Mini High 99.06% 97.19% 100.00% 100.00% 90.29%
4 GPT-5.4 Nano (Reasoning, Low) 99.03% 97.08% 100.00% 100.00% 79.48%
5 Grok 4.1 Fast 99.02% 97.08% 100.00% 99.99% 89.55%
6 o4 Mini 98.75% 96.25% 100.00% 100.00% 88.35%
7 Z.AI GLM 5 Turbo 98.64% 95.94% 100.00% 99.99% 94.27%
8 GPT-5.1 98.44% 95.31% 100.00% 100.00% 92.54%
9 GPT-5.4 Mini (Reasoning, Low) 98.43% 95.31% 100.00% 99.98% 85.75%
10 Claude Opus 4.6 (Reasoning) 98.13% 94.38% 100.00% 100.00% 95.02%
11 GPT-5.4 Nano (Reasoning) 97.95% 93.85% 100.00% 100.00% 81.36%
12 Z.AI GLM 5 97.74% 93.23% 100.00% 100.00% 91.23%
13 GPT-5 Mini 97.71% 93.12% 100.00% 100.00% 92.62%
14 Z.AI GLM 5.1 97.67% 93.02% 100.00% 100.00% 94.37%
15 Gemma 4 26B (Reasoning) 97.59% 92.81% 100.00% 99.95% 91.49%
16 Grok 4.3 (Reasoning) 97.40% 92.19% 100.00% 100.00% 93.60%
17 Claude Opus 4.7 (Reasoning) 97.39% 92.19% 100.00% 99.98% 93.23%
18 Qwen 3.5 Plus (2026-04-20) 97.13% 92.50% 100.00% 98.89% 91.51%
19 MiniMax M2.7 96.47% 89.69% 100.00% 99.72% 89.10%
20 Grok 4.20 (Beta, Reasoning) 96.28% 88.85% 100.00% 100.00% 91.49%
21 GPT-5.4 Mini (Reasoning) 96.22% 88.65% 100.00% 100.00% 90.65%
22 Qwen 3.6 27B 95.73% 87.24% 100.00% 99.95% 89.72%
23 MiniMax M3 95.66% 86.98% 100.00% 100.00% 90.88%
24 Grok 4.20 (Reasoning) 95.66% 86.98% 100.00% 100.00% 91.39%
25 Gemini 2.5 Flash (Reasoning) 95.60% 86.87% 100.00% 99.94% 86.51%
26 Gemini 2.5 Flash Lite (Reasoning) 95.59% 91.56% 95.24% 99.96% 85.75%
27 GPT-OSS 120B 95.29% 85.94% 100.00% 99.92% 86.44%
28 Xiaomi MIMO v2.5 95.25% 85.83% 100.00% 99.93% 85.05%
29 GPT-4.1 95.24% 85.76% 100.00% 99.97% 88.68%
30 Xiaomi MIMO v2.5 Pro 95.17% 85.52% 100.00% 100.00% 87.36%
31 ByteDance Seed 1.6 95.14% 85.42% 100.00% 100.00% 90.70%
32 GPT-5.2 95.09% 85.28% 100.00% 100.00% 90.26%
33 Inception Mercury 95.08% 98.13% 92.86% 94.25% 79.50%
34 DeepSeek V4 Flash (Reasoning) 95.02% 88.65% 96.43% 100.00% 89.01%
35 Claude Opus 4.8 (Reasoning) 94.83% 84.48% 100.00% 100.00% 92.22%
36 Stealth: Healer Alpha 94.67% 84.06% 100.00% 99.93% 85.93%
37 Claude Opus 4.8 (Reasoning, Low) 94.65% 83.96% 100.00% 100.00% 92.14%
38 ByteDance Seed 1.6 Flash 94.52% 94.11% 89.88% 99.57% 73.27%
39 Gemma 4 31B (Reasoning) 94.41% 83.23% 100.00% 100.00% 91.71%
40 Qwen3.7 Max 94.32% 83.02% 100.00% 99.95% 95.75%
41 Claude Sonnet 4.6 (Reasoning) 93.96% 81.88% 100.00% 100.00% 93.66%
42 Claude Opus 4.6 93.60% 80.81% 100.00% 100.00% 92.35%
43 Aion 2.0 93.57% 87.81% 92.89% 100.00% 89.21%
44 GPT-5 Nano 93.52% 92.50% 88.10% 99.96% 82.60%
45 Mistral Small 4 (Reasoning) 92.98% 82.50% 96.43% 100.00% 82.39%
46 MiniMax M2.5 92.94% 79.32% 100.00% 99.50% 88.71%
47 Z.AI GLM 4.5 Air 92.74% 79.34% 100.00% 98.87% 83.12%
48 Inception Mercury 2 92.60% 96.25% 81.55% 99.99% 83.85%
49 Ministral 3 8B 92.52% 77.55% 100.00% 100.00% 71.76%
50 GPT-5.4 (Reasoning, Low) 92.29% 76.88% 100.00% 100.00% 91.41%
51 Qwen 3.6 Flash 92.26% 76.77% 100.00% 100.00% 90.65%
52 GPT-5 91.84% 75.52% 100.00% 100.00% 91.93%
53 Qwen3.6 Max Preview 91.72% 75.16% 100.00% 100.00% 94.54%
54 ByteDance Seed 2.0 Mini 91.29% 73.98% 100.00% 99.90% 86.91%
55 Gemini 3.5 Flash (Reasoning) 91.22% 73.65% 100.00% 100.00% 94.08%
56 Arcee AI: Trinity Mini 91.12% 73.37% 100.00% 100.00% 70.90%
57 Grok 4 Fast 91.09% 73.28% 100.00% 100.00% 86.15%
58 DeepSeek V4 Pro (Reasoning) 90.88% 73.75% 100.00% 98.89% 90.10%
59 Stealth: Hunter Alpha 90.78% 72.33% 100.00% 100.00% 87.34%
60 LFM2 24B 90.53% 71.58% 100.00% 100.00% 58.77%
61 GPT-5.4 (Reasoning) 90.43% 71.30% 100.00% 100.00% 93.24%
62 Qwen 3.6 35B 90.19% 71.67% 100.00% 98.89% 89.05%
63 Gemma 4 31B 90.17% 70.52% 100.00% 100.00% 86.91%
64 Z.AI GLM 4.6 90.08% 71.35% 100.00% 98.89% 89.11%
65 Claude Sonnet 4.6 89.98% 69.98% 100.00% 99.97% 91.15%
66 Z.AI GLM 4.7 Flash 89.86% 76.72% 92.86% 100.00% 84.82%
67 Qwen 3 32B 89.56% 69.27% 100.00% 99.39% 82.21%
68 Grok 4 89.45% 68.35% 100.00% 100.00% 88.12%
69 GPT-5.4 Nano 89.29% 67.86% 100.00% 100.00% 74.40%
70 Ministral 8B 89.19% 67.57% 100.00% 100.00% 64.87%
71 Nemotron 3 Nano 89.16% 87.60% 82.14% 97.72% 77.73%
72 Gemini 3.1 Pro (Preview) 89.06% 67.19% 100.00% 100.00% 94.37%
73 MoonshotAI: Kimi K2.6 88.85% 66.56% 100.00% 100.00% 92.31%
74 Z.AI GLM 4.7 88.47% 65.47% 100.00% 99.96% 88.69%
75 Gemma 4 26B 88.45% 65.34% 100.00% 100.00% 85.84%
76 Gemini 3 Pro (Preview) 88.23% 64.69% 100.00% 100.00% 88.79%
77 MoonshotAI: Kimi K2.5 88.02% 64.06% 100.00% 100.00% 91.04%
78 GPT-4.1 Nano 87.73% 63.25% 100.00% 99.95% 71.94%
79 Z.AI GLM 4.5 87.05% 62.31% 100.00% 98.85% 86.27%
80 Qwen 3.5 122B 86.58% 59.74% 100.00% 100.00% 91.53%
81 Qwen 3.5 27B 86.58% 59.73% 100.00% 100.00% 90.85%
82 Gemini 2.5 Pro 86.11% 58.32% 100.00% 100.00% 88.53%
83 Claude Opus 4.7 85.89% 57.68% 100.00% 100.00% 89.93%
84 Qwen 3.5 9B 85.58% 62.64% 96.43% 97.67% 86.05%
85 Gemini 3 Flash (Preview, Reasoning) 85.32% 55.98% 100.00% 99.97% 90.50%
86 GPT-5.5 (Reasoning, Low) 84.41% 53.23% 100.00% 100.00% 92.59%
87 GPT-5.5 (Reasoning) 84.24% 52.71% 100.00% 100.00% 92.98%
88 Claude Haiku 4.5 83.86% 51.59% 100.00% 100.00% 85.14%
89 Qwen 3.5 397B A17B 82.10% 46.35% 100.00% 99.93% 91.73%
90 Mistral Medium 3.1 82.09% 46.27% 100.00% 100.00% 77.83%
91 Claude Opus 4.5 82.06% 46.19% 100.00% 100.00% 89.69%
92 Cydonia 24B V4.1 81.53% 55.90% 91.07% 97.62% 75.09%
93 GPT-4.1 Mini 81.14% 43.42% 100.00% 100.00% 83.20%
94 Qwen 3.5 35B 80.87% 49.74% 92.86% 100.00% 88.00%
95 Qwen 3.5 Flash 80.63% 49.01% 92.90% 99.98% 86.38%
96 GPT-5.5 80.57% 41.70% 100.00% 100.00% 89.09%
97 Claude Sonnet 4 80.08% 40.23% 100.00% 100.00% 88.72%
98 Ministral 3 14B 79.99% 39.97% 100.00% 100.00% 72.54%
99 ByteDance Seed 2.0 Lite 79.75% 41.46% 100.00% 97.78% 84.80%
100 Hermes 3 405B 79.70% 39.09% 100.00% 100.00% 82.86%
101 Qwen 2.5 72B 79.56% 40.41% 100.00% 98.28% 75.46%
102 GPT-4o, Aug. 6th (temp=1) 79.53% 38.76% 100.00% 99.82% 82.62%
103 Gemini 3.5 Flash (Reasoning, Minimal) 79.09% 37.28% 100.00% 100.00% 86.47%
104 DeepSeek V4 Pro 78.72% 36.18% 100.00% 99.98% 82.63%
105 GPT-5.4 Mini 78.40% 35.21% 100.00% 100.00% 82.43%
106 Grok 4.20 (Beta) 78.28% 34.84% 100.00% 100.00% 83.85%
107 Mistral Large 3 78.17% 34.52% 100.00% 100.00% 85.43%
108 Mistral Large 2 77.87% 33.60% 100.00% 100.00% 82.41%
109 Mistral Large 77.50% 32.50% 100.00% 100.00% 80.15%
110 Grok 4.3 77.16% 31.51% 100.00% 99.96% 78.66%
111 GPT-4o Mini (temp=1) 76.78% 30.35% 100.00% 100.00% 79.08%
112 Arcee AI: Trinity Large (Preview) 76.47% 29.41% 100.00% 100.00% 73.33%
113 Claude 3.5 Sonnet 76.31% 28.94% 100.00% 100.00% 84.24%
114 Gemini 2.5 Flash Lite 76.17% 28.56% 100.00% 99.96% 81.08%
115 Gemini 3.1 Flash Lite 75.72% 27.15% 100.00% 100.00% 85.75%
116 Claude Opus 4 75.68% 27.04% 100.00% 100.00% 87.69%
117 Mistral Small 3.2 24B 75.64% 28.78% 100.00% 98.14% 78.58%
118 Claude Sonnet 4.5 75.57% 26.70% 100.00% 100.00% 88.03%
119 DeepSeek V4 Flash 75.47% 26.40% 100.00% 100.00% 82.02%
120 Gemini 3.1 Flash Lite (Reasoning) 75.26% 25.79% 100.00% 100.00% 86.41%
121 Claude 3.7 Sonnet 75.18% 25.54% 100.00% 100.00% 83.39%
122 Llama 3.1 Nemotron 70B 74.99% 28.79% 96.43% 99.75% 74.70%
123 Gemini 3.1 Flash Lite (Preview) 74.58% 23.78% 100.00% 99.97% 85.87%
124 Mistral Small Creative 74.46% 23.38% 100.00% 100.00% 73.27%
125 GPT-5.4 73.78% 21.34% 100.00% 100.00% 84.32%
126 GPT-4o Mini (temp=0) 73.56% 20.67% 100.00% 100.00% 78.29%
127 Mistral Small 4 73.41% 20.22% 100.00% 100.00% 76.46%
128 GPT-4o, May 13th (temp=1) 73.40% 20.21% 100.00% 100.00% 83.80%
129 GPT-4o, Aug. 6th (temp=0) 73.35% 20.06% 100.00% 100.00% 82.45%
130 Qwen 3.5 Plus (2026-02-15) 73.35% 20.40% 100.00% 99.65% 85.96%
131 DeepSeek V3 (2024-12-26) 73.11% 19.33% 100.00% 100.00% 83.68%
132 Grok 4.20 73.01% 19.05% 100.00% 99.99% 81.70%
133 DeepSeek V3.1 72.80% 18.52% 100.00% 99.89% 82.39%
134 DeepSeek V3.2 72.50% 17.51% 100.00% 100.00% 82.25%
135 Writer: Palmyra X5 72.04% 23.27% 92.86% 100.00% 79.57%
136 Gemini 2.5 Flash 71.70% 16.03% 100.00% 99.08% 80.60%
137 Gemini 3 Flash (Preview) 71.24% 13.77% 100.00% 99.95% 85.35%
138 Ministral 3B 70.75% 26.52% 85.71% 100.00% 61.29%
139 Ministral 3 3B 70.45% 22.08% 89.29% 100.00% 67.22%
140 WizardLM 2 8x22b 70.19% 29.74% 85.73% 95.09% 71.06%
141 Qwen3 235B A22B Instruct 2507 69.82% 23.76% 85.71% 99.97% 80.10%
142 Llama 3.1 70B 69.78% 29.26% 92.86% 87.24% 78.40%
143 GPT-4o, May 13th (temp=0) 69.76% 9.29% 100.00% 100.00% 85.36%
144 DeepSeek-V2 Chat 69.48% 17.97% 90.48% 100.00% 84.83%
145 Gemma 3 12B 69.15% 7.45% 100.00% 100.00% 78.41%
146 Gemma 3 27B 68.74% 6.23% 100.00% 100.00% 77.85%
147 Gemma 3 4B 67.60% 2.80% 100.00% 100.00% 68.57%
148 DeepSeek V3 (2025-03-24) 67.07% 18.55% 88.10% 94.58% 81.99%
149 Hermes 3 70B 67.03% 42.95% 60.79% 97.34% 72.57%
150 Skyfall 36B V2 66.29% 44.87% 56.07% 97.93% 65.76%
151 Cohere Command R+ (Aug. 2024) 64.40% 55.29% 37.90% 100.00% 69.03%
152 Mistral NeMO 62.63% 11.73% 76.19% 99.97% 65.04%
153 Claude 3 Haiku 60.81% 19.94% 62.50% 100.00% 71.19%
154 Rocinante 12B 52.07% 27.40% 38.95% 89.87% 54.54%
155 Llama 3.1 8B 43.27% 18.43% 77.38% 34.01% 63.35%