Matches Regex

Test: Data extraction

Avg. Score
99.0%
Scenarios
9

Overall Performance

Rank ▲ Model Score Avg. Cost Avg. Time Stability
1Gemma 3 4B100.0%$0.0000249ms100%
2Gemini 2.5 Flash Lite100.0%$0.0000355ms100%
3Gemma 3 12B100.0%$0.0000414ms100%
4Ministral 3 14B100.0%$0.0000424ms100%
5Mistral Small 3.2 24B100.0%$0.0000505ms100%
6Gemini 2.5 Flash100.0%$0.0000455ms100%
7Inception Mercury 2100.0%$0.0001361ms100%
8Gemma 3 27B100.0%$0.0000691ms100%
9Llama 3.1 70B100.0%$0.0001533ms100%
10Mistral Medium 3.1100.0%$0.0000650ms100%
11Gemini 3.1 Flash Lite (Preview)100.0%$0.0000697ms100%
12Gemini 3.1 Flash Lite100.0%$0.0000729ms100%
13Qwen 2.5 72B100.0%$0.0000715ms100%
14Hermes 3 70B100.0%$0.0000742ms100%
15GPT-5.4 Mini100.0%$0.0001662ms100%
16Mistral Large 2100.0%$0.0002436ms100%
17Gemini 3 Flash (Preview)100.0%$0.0000831ms100%
18DeepSeek V3 (2024-12-26)100.0%$0.0000897ms100%
19GPT-4.1 Nano100.0%$0.0000994ms100%
20Mistral Large 3100.0%$0.00001.0s100%
21Gemini 3.1 Flash Lite (Reasoning)100.0%$0.00001.2s100%
22GPT-5.4100.0%$0.0003652ms100%
23Claude Haiku 4.5100.0%$0.00011.1s100%
24GPT-4.1 Mini100.0%$0.00001.4s100%
25DeepSeek-V2 Chat100.0%$0.00001.5s100%
26GPT-4o, Aug. 6th (temp=1)100.0%$0.0002962ms100%
27GPT-4o, Aug. 6th (temp=0)100.0%$0.0002966ms100%
28GPT-5.4 Nano (Reasoning, Low)100.0%$0.00001.5s100%
29GPT-4.1100.0%$0.00021.2s100%
30DeepSeek V3 (2025-03-24)100.0%$0.00001.7s100%
31Gemma 4 26B100.0%$0.00001.9s100%
32ByteDance Seed 1.6 Flash100.0%$0.00011.8s100%
33GPT-5.4 Nano (Reasoning)100.0%$0.00011.9s100%
34Gemini 2.5 Flash Lite (Reasoning)100.0%$0.00011.9s100%
35Claude Sonnet 4.6100.0%$0.00041.1s100%
36GPT-5.4 Mini (Reasoning, Low)100.0%$0.00021.9s100%
37Claude Sonnet 4100.0%$0.00041.5s100%
38GPT-5.5100.0%$0.00051.2s100%
39Claude Sonnet 4.5100.0%$0.00031.9s100%
40Z.AI GLM 4.5100.0%$0.00012.7s100%
41GPT-5.4 Mini (Reasoning)100.0%$0.00032.1s100%
42Claude Opus 4.7100.0%$0.0007978ms100%
43Claude Opus 4.7 (Reasoning)100.0%$0.0007988ms100%
44Nemotron 3 Super100.0%$0.00003.2s100%
45GPT-5.4 (Reasoning, Low)100.0%$0.00061.8s100%
46DeepSeek V4 Flash (Reasoning)100.0%$0.00003.4s100%
47Claude Sonnet 5100.0%$0.00032.7s100%
48GPT-5.2100.0%$0.00062.1s100%
49GPT-5 Mini100.0%$0.00033.1s100%
50Claude Sonnet 5 (Reasoning)100.0%$0.00033.0s100%
51Claude Opus 4.6100.0%$0.00062.4s100%
52GPT-5.1100.0%$0.00052.7s100%
53MiniMax M3100.0%$0.00013.8s100%
54Claude Sonnet 5 (Reasoning, Low)100.0%$0.00033.3s100%
55Claude Sonnet 4.6 (Reasoning)100.0%$0.00101.6s100%
56Claude Opus 4.5100.0%$0.00092.2s100%
57ByteDance Seed 2.0 Lite100.0%$0.00034.0s100%
58o4 Mini High100.0%$0.00073.0s100%
59Qwen 3.6 Flash100.0%$0.00073.0s100%
60GPT-5.4 (Reasoning)100.0%$0.00092.6s100%
61GPT-OSS 120B100.0%$0.00005.3s100%
62Gemini 3 Flash (Preview, Reasoning)100.0%$0.00063.8s100%
63Gemma 4 31B100.0%$0.00005.5s100%
64Z.AI GLM 5 Turbo100.0%$0.00083.4s100%
65GPT-5 Nano100.0%$0.00015.3s100%
66Qwen 3.6 35B100.0%$0.00064.1s100%
67Writer: Palmyra X5100.0%$0.00025.2s100%
68Qwen 3.5 Flash100.0%$0.00025.2s100%
69Mistral Small 498.9%$0.0000509ms79%
70Claude Opus 4.8 (Reasoning, Low)100.0%$0.00122.4s100%
71Xiaomi MIMO v2.5100.0%$0.00074.0s100%
72Claude Opus 4.8 (Reasoning)100.0%$0.00122.6s100%
73GPT-5.4 Nano98.9%$0.0000711ms79%
74WizardLM 2 8x22b100.0%$0.00016.1s100%
75ByteDance Seed 1.6100.0%$0.00045.3s100%
76GPT-5.5 (Reasoning, Low)100.0%$0.00132.7s100%
77Gemini 3.5 Flash (Reasoning, Minimal)98.9%$0.0001851ms79%
78Qwen3 235B A22B Instruct 250798.9%$0.00001.4s79%
79Nemotron 3 Nano100.0%$0.00006.7s100%
80Qwen 3 32B100.0%$0.00016.5s100%
81MiniMax M2.7100.0%$0.00035.9s100%
82GPT-5.5 (Reasoning)100.0%$0.00152.7s100%
83Claude Opus 4.6 (Reasoning)100.0%$0.00172.5s100%
84MiniMax M2.5100.0%$0.00046.5s100%
85GPT-4o Mini (temp=0)100.0%$0.00007.6s100%
86DeepSeek V3.198.9%$0.00002.3s79%
87Qwen 3.5 Plus (2026-02-15)98.9%$0.00002.4s79%
88Qwen 3.5 35B100.0%$0.00134.3s100%
89Z.AI GLM 5.2 (Reasoning, High)100.0%$0.00115.4s100%
90Gemma 4 26B (Reasoning)100.0%$0.00018.4s100%
91Gemini 3.5 Flash (Reasoning)100.0%$0.00241.8s100%
92DeepSeek V4 Pro (Reasoning)100.0%$0.00038.0s100%
93Xiaomi MIMO v2.5 Pro100.0%$0.00106.3s100%
94ByteDance Seed 2.0 Mini100.0%$0.00018.8s100%
95Z.AI GLM 4.7 Flash100.0%$0.00028.7s100%
96MoonshotAI: Kimi K2.5100.0%$0.00087.1s100%
97GPT-5100.0%$0.00184.3s100%
98Aion 2.0100.0%$0.00058.0s100%
99Grok 4.20 (Reasoning)100.0%$0.00146.1s100%
100Grok 4.5 (Reasoning, High)100.0%$0.00165.9s100%
101Mistral Small 4 (Reasoning)97.8%$0.00012.1s71%
102Grok 4.5 (Reasoning, Low)100.0%$0.00136.9s100%
103Qwen 3.5 27B100.0%$0.00146.8s100%
104Mistral NeMO97.8%$0.00002.9s71%
105Claude Opus 4100.0%$0.00176.2s100%
106GPT-4o Mini (temp=1)100.0%$0.000011.3s100%
107MoonshotAI: Kimi K2.6100.0%$0.00098.7s100%
108DeepSeek V3.297.8%$0.00003.7s71%
109Arcee AI: Trinity Mini96.7%$0.00012.1s64%
110Gemini 2.5 Flash (Reasoning)97.8%$0.00082.5s71%
111Qwen 3.5 Plus (2026-04-20)100.0%$0.00129.7s100%
112Z.AI GLM 5.1100.0%$0.001210.0s100%
113Hermes 3 405B97.8%$0.00006.3s71%
114Qwen 3.5 9B100.0%$0.000114.4s100%
115Aion 3.0 Mini98.9%$0.00058.2s79%
116Gemini 3.1 Pro (Preview)100.0%$0.00325.9s100%
117Grok 4.3 (Reasoning)100.0%$0.001212.0s100%
118Gemini 2.5 Pro100.0%$0.00404.9s100%
119Aion 3.097.8%$0.00134.6s71%
120Cydonia 24B V4.193.3%$0.0000531ms50%
121Grok 4.393.3%$0.0002727ms50%
122Qwen3.7 Max100.0%$0.00348.1s100%
123Qwen 3.6 27B98.9%$0.00158.3s79%
124DeepSeek V4 Pro94.4%$0.00014.2s54%
125Gemma 4 31B (Reasoning)100.0%$0.000118.6s100%
126Z.AI GLM 4.5 Air98.9%$0.000512.1s79%
127Grok 4.2092.2%$0.00021.1s46%
128o4 Mini100.0%$0.000617.7s100%
129Z.AI GLM 4.7100.0%$0.000817.4s100%
130Z.AI GLM 5100.0%$0.001515.9s100%
131Ministral 8B90.0%$0.0000307ms40%
132Z.AI GLM 4.6100.0%$0.001118.3s100%
133Qwen 3.5 397B A17B100.0%$0.002115.7s100%
134Ministral 3 3B88.9%$0.0000365ms37%
135Cohere Command R+ (Aug. 2024)88.9%$0.0002446ms37%
136Qwen3.6 Max Preview100.0%$0.003514.0s100%
137Ministral 3B85.6%$0.0000297ms30%
138Qwen 3.5 122B100.0%$0.006315.7s100%
139Ministral 3 8B77.8%$0.0000372ms17%
140DeepSeek V4 Flash76.7%$0.00002.9s15%
98.96%

Individual Scenarios

Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 # 8 # 9 # 10 Avg ▼
Claude Opus 4.6 (Reasoning)100100100100100100100100100100100.0%
Qwen3.7 Max100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100100100100.0%
Qwen3.6 Max Preview100100100100100100100100100100100.0%
GPT-5.5 (Reasoning)100100100100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.6100100100100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5100100100100100100100100100100100.0%
GPT-5 Mini100100100100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100100100100.0%
Grok 4.3 (Reasoning)100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.5100100100100100100100100100100100.0%
GPT-5.1100100100100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100100100100.0%
MiniMax M3100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100100100100.0%
GPT-5.2100100100100100100100100100100100.0%
GPT-5.5100100100100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100100100100.0%
o4 Mini High100100100100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100100100100.0%
Claude Opus 4100100100100100100100100100100100.0%
Qwen 3.5 35B100100100100100100100100100100100.0%
GPT-4.1100100100100100100100100100100100.0%
MiniMax M2.5100100100100100100100100100100100.0%
Aion 2.0100100100100100100100100100100100.0%
o4 Mini100100100100100100100100100100100.0%
MiniMax M2.7100100100100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100100100100.0%
Qwen 3.5 Flash100100100100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100100100100.0%
Gemma 4 31B100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100100100100.0%
Z.AI GLM 4.5100100100100100100100100100100100.0%
Gemma 4 26B100100100100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100100100100.0%
GPT-5.4100100100100100100100100100100100.0%
Mistral Large 3100100100100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100100100100.0%
Gemini 2.5 Flash (Reasoning)100100100100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100100100100.0%
Qwen 3.5 9B100100100100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100100100100.0%
Aion 3.0 Mini100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)100100100100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100100100100.0%
DeepSeek V3.1100100100100100100100100100100100.0%
DeepSeek V3.2100100100100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100100100100.0%
DeepSeek V4 Pro100100100100100100100100100100100.0%
Inception Mercury 2100100100100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100100100100.0%
Mistral Large 2100100100100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100100100100.0%
Grok 4.20100100100100100100100100100100100.0%
Hermes 3 405B100100100100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100100100100.0%
GPT-5 Nano100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100100100100.0%
Mistral Small 4 (Reasoning)100100100100100100100100100100100.0%
Qwen 3 32B100100100100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100100100100.0%
Qwen3 235B A22B Instruct 2507100100100100100100100100100100100.0%
Grok 4.3100100100100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100100100100.0%
Gemma 3 12B100100100100100100100100100100100.0%
Gemma 3 27B100100100100100100100100100100100.0%
Mistral Small 4100100100100100100100100100100100.0%
Nemotron 3 Nano100100100100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100100100100.0%
GPT-5.4 Nano100100100100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100100100100.0%
Ministral 3 14B100100100100100100100100100100100.0%
Ministral 3 8B100100100100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100100100100.0%
Hermes 3 70B100100100100100100100100100100100.0%
Arcee AI: Trinity Mini100100100100100100100100100100100.0%
Cohere Command R+ (Aug. 2024)100100100100100100100100100100100.0%
Gemma 3 4B100100100100100100100100100100100.0%
Ministral 3 3B100100100100100100100100100100100.0%
Mistral NeMO100100100100100100100100100100100.0%
Aion 3.0100100100100100100100100100090.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100100090.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100100090.0%
DeepSeek V4 Flash100100100100100100100100100090.0%
Cydonia 24B V4.1100100100100100100100100100090.0%
Ministral 8B100100100100100100100100100090.0%
Ministral 3B1001001001001001001001000080.0%
Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 # 8 # 9 # 10 Avg ▼
Claude Opus 4.6 (Reasoning)100100100100100100100100100100100.0%
Qwen3.7 Max100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100100100100.0%
Qwen3.6 Max Preview100100100100100100100100100100100.0%
GPT-5.5 (Reasoning)100100100100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.6100100100100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5100100100100100100100100100100100.0%
GPT-5 Mini100100100100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100100100100.0%
Grok 4.3 (Reasoning)100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.5100100100100100100100100100100100.0%
GPT-5.1100100100100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100100100100.0%
MiniMax M3100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100100100100.0%
GPT-5.2100100100100100100100100100100100.0%
GPT-5.5100100100100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100100100100.0%
Aion 3.0100100100100100100100100100100100.0%
o4 Mini High100100100100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100100100100.0%
Claude Opus 4100100100100100100100100100100100.0%
Qwen 3.5 35B100100100100100100100100100100100.0%
GPT-4.1100100100100100100100100100100100.0%
MiniMax M2.5100100100100100100100100100100100.0%
Aion 2.0100100100100100100100100100100100.0%
o4 Mini100100100100100100100100100100100.0%
MiniMax M2.7100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100100100100.0%
Qwen 3.5 Flash100100100100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100100100100.0%
Gemma 4 31B100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100100100100.0%
Z.AI GLM 4.5100100100100100100100100100100100.0%
Gemma 4 26B100100100100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100100100100.0%
GPT-5.4100100100100100100100100100100100.0%
Mistral Large 3100100100100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100100100100.0%
Gemini 2.5 Flash (Reasoning)100100100100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100100100100.0%
Qwen 3.5 9B100100100100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100100100100.0%
Aion 3.0 Mini100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)100100100100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100100100100.0%
DeepSeek V3.1100100100100100100100100100100100.0%
DeepSeek V3.2100100100100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100100100100.0%
DeepSeek V4 Pro100100100100100100100100100100100.0%
DeepSeek V4 Flash100100100100100100100100100100100.0%
Inception Mercury 2100100100100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100100100100.0%
Mistral Large 2100100100100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100100100100.0%
Grok 4.20100100100100100100100100100100100.0%
Hermes 3 405B100100100100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100100100100.0%
GPT-5 Nano100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100100100100.0%
Mistral Small 4 (Reasoning)100100100100100100100100100100100.0%
Qwen 3 32B100100100100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100100100100.0%
Qwen3 235B A22B Instruct 2507100100100100100100100100100100100.0%
Grok 4.3100100100100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100100100100.0%
Gemma 3 12B100100100100100100100100100100100.0%
Gemma 3 27B100100100100100100100100100100100.0%
Mistral Small 4100100100100100100100100100100100.0%
Nemotron 3 Nano100100100100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100100100100.0%
Cydonia 24B V4.1100100100100100100100100100100100.0%
GPT-5.4 Nano100100100100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100100100100.0%
Ministral 3 14B100100100100100100100100100100100.0%
Ministral 3 8B100100100100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100100100100.0%
Hermes 3 70B100100100100100100100100100100100.0%
Arcee AI: Trinity Mini100100100100100100100100100100100.0%
Cohere Command R+ (Aug. 2024)100100100100100100100100100100100.0%
Gemma 3 4B100100100100100100100100100100100.0%
Ministral 3 3B100100100100100100100100100100100.0%
Mistral NeMO100100100100100100100100100100100.0%
Ministral 8B100100100100100100100100100100100.0%
Ministral 3B100100100100100100100100100100100.0%
Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 # 8 # 9 # 10 Avg ▼
Claude Opus 4.6 (Reasoning)100100100100100100100100100100100.0%
Qwen3.7 Max100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100100100100.0%
Qwen3.6 Max Preview100100100100100100100100100100100.0%
GPT-5.5 (Reasoning)100100100100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.6100100100100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5100100100100100100100100100100100.0%
GPT-5 Mini100100100100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100100100100.0%
Grok 4.3 (Reasoning)100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.5100100100100100100100100100100100.0%
GPT-5.1100100100100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100100100100.0%
MiniMax M3100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100100100100.0%
GPT-5.2100100100100100100100100100100100.0%
GPT-5.5100100100100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100100100100.0%
Aion 3.0100100100100100100100100100100100.0%
o4 Mini High100100100100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100100100100.0%
Claude Opus 4100100100100100100100100100100100.0%
Qwen 3.5 35B100100100100100100100100100100100.0%
GPT-4.1100100100100100100100100100100100.0%
MiniMax M2.5100100100100100100100100100100100.0%
Aion 2.0100100100100100100100100100100100.0%
o4 Mini100100100100100100100100100100100.0%
MiniMax M2.7100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100100100100.0%
Qwen 3.5 Flash100100100100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100100100100.0%
Gemma 4 31B100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100100100100.0%
Z.AI GLM 4.5100100100100100100100100100100100.0%
Gemma 4 26B100100100100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100100100100.0%
GPT-5.4100100100100100100100100100100100.0%
Mistral Large 3100100100100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100100100100.0%
Gemini 2.5 Flash (Reasoning)100100100100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100100100100.0%
Qwen 3.5 9B100100100100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100100100100.0%
Aion 3.0 Mini100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)100100100100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100100100100.0%
DeepSeek V3.1100100100100100100100100100100100.0%
DeepSeek V3.2100100100100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100100100100.0%
DeepSeek V4 Pro100100100100100100100100100100100.0%
DeepSeek V4 Flash100100100100100100100100100100100.0%
Inception Mercury 2100100100100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100100100100.0%
Mistral Large 2100100100100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100100100100.0%
Grok 4.20100100100100100100100100100100100.0%
Hermes 3 405B100100100100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100100100100.0%
GPT-5 Nano100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100100100100.0%
Mistral Small 4 (Reasoning)100100100100100100100100100100100.0%
Qwen 3 32B100100100100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100100100100.0%
Qwen3 235B A22B Instruct 2507100100100100100100100100100100100.0%
Grok 4.3100100100100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100100100100.0%
Gemma 3 12B100100100100100100100100100100100.0%
Gemma 3 27B100100100100100100100100100100100.0%
Mistral Small 4100100100100100100100100100100100.0%
Nemotron 3 Nano100100100100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100100100100.0%
Cydonia 24B V4.1100100100100100100100100100100100.0%
GPT-5.4 Nano100100100100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100100100100.0%
Ministral 3 14B100100100100100100100100100100100.0%
Ministral 3 8B100100100100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100100100100.0%
Hermes 3 70B100100100100100100100100100100100.0%
Arcee AI: Trinity Mini100100100100100100100100100100100.0%
Cohere Command R+ (Aug. 2024)100100100100100100100100100100100.0%
Gemma 3 4B100100100100100100100100100100100.0%
Ministral 3 3B100100100100100100100100100100100.0%
Mistral NeMO100100100100100100100100100100100.0%
Ministral 8B100100100100100100100100100100100.0%
Ministral 3B100100100100100100100100100100100.0%
Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 # 8 # 9 # 10 Avg ▼
Claude Opus 4.6 (Reasoning)100100100100100100100100100100100.0%
Qwen3.7 Max100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100100100100.0%
Qwen3.6 Max Preview100100100100100100100100100100100.0%
GPT-5.5 (Reasoning)100100100100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.6100100100100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5100100100100100100100100100100100.0%
GPT-5 Mini100100100100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100100100100.0%
Grok 4.3 (Reasoning)100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.5100100100100100100100100100100100.0%
GPT-5.1100100100100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100100100100.0%
MiniMax M3100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100100100100.0%
GPT-5.2100100100100100100100100100100100.0%
GPT-5.5100100100100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100100100100.0%
Aion 3.0100100100100100100100100100100100.0%
o4 Mini High100100100100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100100100100.0%
Claude Opus 4100100100100100100100100100100100.0%
Qwen 3.5 35B100100100100100100100100100100100.0%
GPT-4.1100100100100100100100100100100100.0%
MiniMax M2.5100100100100100100100100100100100.0%
Aion 2.0100100100100100100100100100100100.0%
o4 Mini100100100100100100100100100100100.0%
MiniMax M2.7100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100100100100.0%
Qwen 3.5 Flash100100100100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100100100100.0%
Gemma 4 31B100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100100100100.0%
Z.AI GLM 4.5100100100100100100100100100100100.0%
Gemma 4 26B100100100100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100100100100.0%
GPT-5.4100100100100100100100100100100100.0%
Mistral Large 3100100100100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100100100100.0%
Gemini 2.5 Flash (Reasoning)100100100100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100100100100.0%
Qwen 3.5 9B100100100100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)100100100100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100100100100.0%
DeepSeek V3.1100100100100100100100100100100100.0%
DeepSeek V3.2100100100100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100100100100.0%
DeepSeek V4 Pro100100100100100100100100100100100.0%
DeepSeek V4 Flash100100100100100100100100100100100.0%
Inception Mercury 2100100100100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100100100100.0%
Mistral Large 2100100100100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100100100100.0%
Grok 4.20100100100100100100100100100100100.0%
Hermes 3 405B100100100100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100100100100.0%
GPT-5 Nano100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100100100100.0%
Qwen 3 32B100100100100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100100100100.0%
Qwen3 235B A22B Instruct 2507100100100100100100100100100100100.0%
Grok 4.3100100100100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100100100100.0%
Gemma 3 12B100100100100100100100100100100100.0%
Gemma 3 27B100100100100100100100100100100100.0%
Mistral Small 4100100100100100100100100100100100.0%
Nemotron 3 Nano100100100100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100100100100.0%
Cydonia 24B V4.1100100100100100100100100100100100.0%
GPT-5.4 Nano100100100100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100100100100.0%
Ministral 3 14B100100100100100100100100100100100.0%
Ministral 3 8B100100100100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100100100100.0%
Hermes 3 70B100100100100100100100100100100100.0%
Cohere Command R+ (Aug. 2024)100100100100100100100100100100100.0%
Gemma 3 4B100100100100100100100100100100100.0%
Ministral 3 3B100100100100100100100100100100100.0%
Mistral NeMO100100100100100100100100100100100.0%
Ministral 8B100100100100100100100100100100100.0%
Ministral 3B100100100100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100100090.0%
Aion 3.0 Mini100100100100100100100100100090.0%
Mistral Small 4 (Reasoning)1001001001001001001001000080.0%
Arcee AI: Trinity Mini10010010010010010010000070.0%
Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 # 8 # 9 # 10 Avg ▼
Claude Opus 4.6 (Reasoning)100100100100100100100100100100100.0%
Qwen3.7 Max100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100100100100.0%
Qwen3.6 Max Preview100100100100100100100100100100100.0%
GPT-5.5 (Reasoning)100100100100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.6100100100100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5100100100100100100100100100100100.0%
GPT-5 Mini100100100100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100100100100.0%
Grok 4.3 (Reasoning)100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.5100100100100100100100100100100100.0%
GPT-5.1100100100100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100100100100.0%
MiniMax M3100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100100100100.0%
GPT-5.2100100100100100100100100100100100.0%
GPT-5.5100100100100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100100100100.0%
Aion 3.0100100100100100100100100100100100.0%
o4 Mini High100100100100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100100100100.0%
Claude Opus 4100100100100100100100100100100100.0%
Qwen 3.5 35B100100100100100100100100100100100.0%
GPT-4.1100100100100100100100100100100100.0%
MiniMax M2.5100100100100100100100100100100100.0%
Aion 2.0100100100100100100100100100100100.0%
o4 Mini100100100100100100100100100100100.0%
MiniMax M2.7100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100100100100.0%
Qwen 3.5 Flash100100100100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100100100100.0%
Gemma 4 31B100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100100100100.0%
Z.AI GLM 4.5100100100100100100100100100100100.0%
Gemma 4 26B100100100100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100100100100.0%
GPT-5.4100100100100100100100100100100100.0%
Mistral Large 3100100100100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100100100100.0%
Gemini 2.5 Flash (Reasoning)100100100100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100100100100.0%
Qwen 3.5 9B100100100100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100100100100.0%
Aion 3.0 Mini100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)100100100100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100100100100.0%
DeepSeek V3.1100100100100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100100100100.0%
Inception Mercury 2100100100100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100100100100.0%
Mistral Large 2100100100100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100100100100.0%
Grok 4.20100100100100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100100100100.0%
GPT-5 Nano100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100100100100.0%
Mistral Small 4 (Reasoning)100100100100100100100100100100100.0%
Qwen 3 32B100100100100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100100100100.0%
Qwen3 235B A22B Instruct 2507100100100100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100100100100.0%
Gemma 3 12B100100100100100100100100100100100.0%
Gemma 3 27B100100100100100100100100100100100.0%
Nemotron 3 Nano100100100100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100100100100.0%
Cydonia 24B V4.1100100100100100100100100100100100.0%
GPT-5.4 Nano100100100100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100100100100.0%
Ministral 3 14B100100100100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100100100100.0%
Hermes 3 70B100100100100100100100100100100100.0%
Arcee AI: Trinity Mini100100100100100100100100100100100.0%
Gemma 3 4B100100100100100100100100100100100.0%
Ministral 3 3B100100100100100100100100100100100.0%
Mistral NeMO100100100100100100100100100100100.0%
Mistral Small 4100100100100100100100100100090.0%
DeepSeek V3.21001001001001001001001000080.0%
Hermes 3 405B1001001001001001001001000080.0%
Grok 4.31001001001001001001001000080.0%
Ministral 8B1001001001001001001001000080.0%
Ministral 3B10010010010010010010000070.0%
DeepSeek V4 Pro1001001001001000000050.0%
DeepSeek V4 Flash00000000000.0%
Ministral 3 8B00000000000.0%
Cohere Command R+ (Aug. 2024)00000000000.0%
Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 # 8 # 9 # 10 Avg ▼
Claude Opus 4.6 (Reasoning)100100100100100100100100100100100.0%
Qwen3.7 Max100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100100100100.0%
Qwen3.6 Max Preview100100100100100100100100100100100.0%
GPT-5.5 (Reasoning)100100100100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.6100100100100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5100100100100100100100100100100100.0%
GPT-5 Mini100100100100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100100100100.0%
Grok 4.3 (Reasoning)100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.5100100100100100100100100100100100.0%
GPT-5.1100100100100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100100100100.0%
MiniMax M3100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100100100100.0%
GPT-5.2100100100100100100100100100100100.0%
GPT-5.5100100100100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100100100100.0%
Aion 3.0100100100100100100100100100100100.0%
o4 Mini High100100100100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100100100100.0%
Claude Opus 4100100100100100100100100100100100.0%
Qwen 3.5 35B100100100100100100100100100100100.0%
GPT-4.1100100100100100100100100100100100.0%
MiniMax M2.5100100100100100100100100100100100.0%
Aion 2.0100100100100100100100100100100100.0%
o4 Mini100100100100100100100100100100100.0%
MiniMax M2.7100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100100100100.0%
Qwen 3.5 Flash100100100100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100100100100.0%
Gemma 4 31B100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100100100100.0%
Z.AI GLM 4.5100100100100100100100100100100100.0%
Gemma 4 26B100100100100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100100100100.0%
GPT-5.4100100100100100100100100100100100.0%
Mistral Large 3100100100100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100100100100.0%
Gemini 2.5 Flash (Reasoning)100100100100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100100100100.0%
Qwen 3.5 9B100100100100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100100100100.0%
Aion 3.0 Mini100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)100100100100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100100100100.0%
DeepSeek V3.2100100100100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100100100100.0%
DeepSeek V4 Pro100100100100100100100100100100100.0%
Inception Mercury 2100100100100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100100100100.0%
Mistral Large 2100100100100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100100100100.0%
Hermes 3 405B100100100100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100100100100.0%
GPT-5 Nano100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100100100100.0%
Mistral Small 4 (Reasoning)100100100100100100100100100100100.0%
Qwen 3 32B100100100100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100100100100.0%
Qwen3 235B A22B Instruct 2507100100100100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100100100100.0%
Gemma 3 12B100100100100100100100100100100100.0%
Gemma 3 27B100100100100100100100100100100100.0%
Mistral Small 4100100100100100100100100100100100.0%
Nemotron 3 Nano100100100100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100100100100.0%
Ministral 3 14B100100100100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100100100100.0%
Hermes 3 70B100100100100100100100100100100100.0%
Arcee AI: Trinity Mini100100100100100100100100100100100.0%
Cohere Command R+ (Aug. 2024)100100100100100100100100100100100.0%
Gemma 3 4B100100100100100100100100100100100.0%
Mistral NeMO100100100100100100100100100100100.0%
DeepSeek V3.1100100100100100100100100100090.0%
GPT-5.4 Nano100100100100100100100100100090.0%
Grok 4.310010010010010010010000070.0%
Cydonia 24B V4.1100100100100100100000060.0%
Ministral 8B1001001001001000000050.0%
Grok 4.20100100100000000030.0%
Ministral 3B1001000000000020.0%
DeepSeek V4 Flash00000000000.0%
Ministral 3 8B00000000000.0%
Ministral 3 3B00000000000.0%
Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 # 8 # 9 # 10 Avg ▼
Claude Opus 4.6 (Reasoning)100100100100100100100100100100100.0%
Qwen3.7 Max100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100100100100.0%
Qwen3.6 Max Preview100100100100100100100100100100100.0%
GPT-5.5 (Reasoning)100100100100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.6100100100100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5100100100100100100100100100100100.0%
GPT-5 Mini100100100100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100100100100.0%
Grok 4.3 (Reasoning)100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.5100100100100100100100100100100100.0%
GPT-5.1100100100100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100100100100.0%
MiniMax M3100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100100100100.0%
GPT-5.2100100100100100100100100100100100.0%
GPT-5.5100100100100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100100100100.0%
o4 Mini High100100100100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100100100100.0%
Claude Opus 4100100100100100100100100100100100.0%
Qwen 3.5 35B100100100100100100100100100100100.0%
GPT-4.1100100100100100100100100100100100.0%
MiniMax M2.5100100100100100100100100100100100.0%
Aion 2.0100100100100100100100100100100100.0%
o4 Mini100100100100100100100100100100100.0%
MiniMax M2.7100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100100100100.0%
Qwen 3.5 Flash100100100100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100100100100.0%
Gemma 4 31B100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100100100100.0%
Z.AI GLM 4.5100100100100100100100100100100100.0%
Gemma 4 26B100100100100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100100100100.0%
GPT-5.4100100100100100100100100100100100.0%
Mistral Large 3100100100100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100100100100.0%
Qwen 3.5 9B100100100100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100100100100.0%
Aion 3.0 Mini100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)100100100100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100100100100.0%
DeepSeek V3.1100100100100100100100100100100100.0%
DeepSeek V3.2100100100100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100100100100.0%
DeepSeek V4 Pro100100100100100100100100100100100.0%
DeepSeek V4 Flash100100100100100100100100100100100.0%
Inception Mercury 2100100100100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100100100100.0%
Mistral Large 2100100100100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100100100100.0%
Grok 4.20100100100100100100100100100100100.0%
Hermes 3 405B100100100100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100100100100.0%
GPT-5 Nano100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100100100100.0%
Mistral Small 4 (Reasoning)100100100100100100100100100100100.0%
Qwen 3 32B100100100100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100100100100.0%
Qwen3 235B A22B Instruct 2507100100100100100100100100100100100.0%
Grok 4.3100100100100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100100100100.0%
Gemma 3 12B100100100100100100100100100100100.0%
Gemma 3 27B100100100100100100100100100100100.0%
Mistral Small 4100100100100100100100100100100100.0%
Nemotron 3 Nano100100100100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100100100100.0%
Cydonia 24B V4.1100100100100100100100100100100100.0%
GPT-5.4 Nano100100100100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100100100100.0%
Ministral 3 14B100100100100100100100100100100100.0%
Ministral 3 8B100100100100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100100100100.0%
Hermes 3 70B100100100100100100100100100100100.0%
Arcee AI: Trinity Mini100100100100100100100100100100100.0%
Cohere Command R+ (Aug. 2024)100100100100100100100100100100100.0%
Gemma 3 4B100100100100100100100100100100100.0%
Ministral 3 3B100100100100100100100100100100100.0%
Mistral NeMO100100100100100100100100100100100.0%
Ministral 8B100100100100100100100100100100100.0%
Ministral 3B100100100100100100100100100100100.0%
Aion 3.0100100100100100100100100100090.0%
Gemini 2.5 Flash (Reasoning)1001001001001001001001000080.0%
Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 # 8 # 9 # 10 Avg ▼
Claude Opus 4.6 (Reasoning)100100100100100100100100100100100.0%
Qwen3.7 Max100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100100100100.0%
Qwen3.6 Max Preview100100100100100100100100100100100.0%
GPT-5.5 (Reasoning)100100100100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.6100100100100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5100100100100100100100100100100100.0%
GPT-5 Mini100100100100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100100100100.0%
Grok 4.3 (Reasoning)100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.5100100100100100100100100100100100.0%
GPT-5.1100100100100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100100100100.0%
MiniMax M3100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100100100100.0%
GPT-5.2100100100100100100100100100100100.0%
GPT-5.5100100100100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100100100100.0%
Aion 3.0100100100100100100100100100100100.0%
o4 Mini High100100100100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100100100100.0%
Claude Opus 4100100100100100100100100100100100.0%
Qwen 3.5 35B100100100100100100100100100100100.0%
GPT-4.1100100100100100100100100100100100.0%
MiniMax M2.5100100100100100100100100100100100.0%
Aion 2.0100100100100100100100100100100100.0%
o4 Mini100100100100100100100100100100100.0%
MiniMax M2.7100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100100100100.0%
Qwen 3.5 Flash100100100100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100100100100.0%
Gemma 4 31B100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100100100100.0%
Z.AI GLM 4.5100100100100100100100100100100100.0%
Gemma 4 26B100100100100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100100100100.0%
GPT-5.4100100100100100100100100100100100.0%
Mistral Large 3100100100100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100100100100.0%
Gemini 2.5 Flash (Reasoning)100100100100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100100100100.0%
Qwen 3.5 9B100100100100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100100100100.0%
Aion 3.0 Mini100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)100100100100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100100100100.0%
DeepSeek V3.1100100100100100100100100100100100.0%
DeepSeek V3.2100100100100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100100100100.0%
DeepSeek V4 Pro100100100100100100100100100100100.0%
DeepSeek V4 Flash100100100100100100100100100100100.0%
Inception Mercury 2100100100100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100100100100.0%
Mistral Large 2100100100100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100100100100.0%
Grok 4.20100100100100100100100100100100100.0%
Hermes 3 405B100100100100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100100100100.0%
GPT-5 Nano100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100100100100.0%
Mistral Small 4 (Reasoning)100100100100100100100100100100100.0%
Qwen 3 32B100100100100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100100100100.0%
Grok 4.3100100100100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100100100100.0%
Gemma 3 12B100100100100100100100100100100100.0%
Gemma 3 27B100100100100100100100100100100100.0%
Mistral Small 4100100100100100100100100100100100.0%
Nemotron 3 Nano100100100100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100100100100.0%
Cydonia 24B V4.1100100100100100100100100100100100.0%
GPT-5.4 Nano100100100100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100100100100.0%
Ministral 3 14B100100100100100100100100100100100.0%
Ministral 3 8B100100100100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100100100100.0%
Hermes 3 70B100100100100100100100100100100100.0%
Arcee AI: Trinity Mini100100100100100100100100100100100.0%
Cohere Command R+ (Aug. 2024)100100100100100100100100100100100.0%
Gemma 3 4B100100100100100100100100100100100.0%
Ministral 3 3B100100100100100100100100100100100.0%
Mistral NeMO100100100100100100100100100100100.0%
Ministral 8B100100100100100100100100100100100.0%
Ministral 3B100100100100100100100100100100100.0%
Qwen3 235B A22B Instruct 2507100100100100100100100100100090.0%
Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 # 8 # 9 # 10 Avg ▼
Claude Opus 4.6 (Reasoning)100100100100100100100100100100100.0%
Qwen3.7 Max100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100100100100.0%
Qwen3.6 Max Preview100100100100100100100100100100100.0%
GPT-5.5 (Reasoning)100100100100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.6100100100100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5100100100100100100100100100100100.0%
GPT-5 Mini100100100100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100100100100.0%
Grok 4.3 (Reasoning)100100100100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100100100100.0%
MoonshotAI: Kimi K2.5100100100100100100100100100100100.0%
GPT-5.1100100100100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100100100100.0%
MiniMax M3100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100100100100.0%
GPT-5.2100100100100100100100100100100100.0%
GPT-5.5100100100100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100100100100.0%
Aion 3.0100100100100100100100100100100100.0%
o4 Mini High100100100100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100100100100.0%
Claude Opus 4100100100100100100100100100100100.0%
Qwen 3.5 35B100100100100100100100100100100100.0%
GPT-4.1100100100100100100100100100100100.0%
MiniMax M2.5100100100100100100100100100100100.0%
Aion 2.0100100100100100100100100100100100.0%
o4 Mini100100100100100100100100100100100.0%
MiniMax M2.7100100100100100100100100100100100.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100100100100.0%
Qwen 3.5 Flash100100100100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100100100100.0%
Gemma 4 31B100100100100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100100100100.0%
Z.AI GLM 4.5100100100100100100100100100100100.0%
Gemma 4 26B100100100100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100100100100.0%
GPT-5.4100100100100100100100100100100100.0%
Mistral Large 3100100100100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100100100100.0%
Gemini 2.5 Flash (Reasoning)100100100100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100100100100.0%
Qwen 3.5 9B100100100100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100100100100.0%
Aion 3.0 Mini100100100100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)100100100100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100100100100.0%
DeepSeek V3.1100100100100100100100100100100100.0%
DeepSeek V3.2100100100100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100100100100.0%
DeepSeek V4 Pro100100100100100100100100100100100.0%
DeepSeek V4 Flash100100100100100100100100100100100.0%
Inception Mercury 2100100100100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100100100100.0%
Mistral Large 2100100100100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100100100100.0%
Grok 4.20100100100100100100100100100100100.0%
Hermes 3 405B100100100100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100100100100.0%
GPT-5 Nano100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100100100100.0%
Mistral Small 4 (Reasoning)100100100100100100100100100100100.0%
Qwen 3 32B100100100100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100100100100.0%
Qwen3 235B A22B Instruct 2507100100100100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100100100100.0%
Gemma 3 12B100100100100100100100100100100100.0%
Gemma 3 27B100100100100100100100100100100100.0%
Mistral Small 4100100100100100100100100100100100.0%
Nemotron 3 Nano100100100100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100100100100.0%
GPT-5.4 Nano100100100100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100100100100.0%
Ministral 3 14B100100100100100100100100100100100.0%
Ministral 3 8B100100100100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100100100100.0%
Hermes 3 70B100100100100100100100100100100100.0%
Arcee AI: Trinity Mini100100100100100100100100100100100.0%
Cohere Command R+ (Aug. 2024)100100100100100100100100100100100.0%
Gemma 3 4B100100100100100100100100100100100.0%
Ministral 3 3B100100100100100100100100100100100.0%
Ministral 3B100100100100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100100090.0%
Grok 4.3100100100100100100100100100090.0%
Cydonia 24B V4.1100100100100100100100100100090.0%
Ministral 8B100100100100100100100100100090.0%
Mistral NeMO1001001001001001001001000080.0%