Possessive traps preserved

Test: Text Replacement

Avg. Score
98.4%
Scenarios
2

Overall Performance

Rank ▲ Model Score Avg. Cost Avg. Time Stability
1Gemini 2.5 Flash Lite100.0%$0.00021.3s100%
2Mistral NeMO100.0%$0.00012.2s100%
3Ministral 8B100.0%$0.00012.4s100%
4Ministral 3 8B100.0%$0.00013.1s100%
5GPT-4.1 Nano100.0%$0.00023.0s100%
6Mistral Small 4100.0%$0.00032.5s100%
7Ministral 3 14B100.0%$0.00023.2s100%
8Mistral Small 3.2 24B100.0%$0.00023.6s100%
9Gemini 3.1 Flash Lite (Preview)100.0%$0.00071.6s100%
10Gemini 3.1 Flash Lite (Reasoning)100.0%$0.00071.6s100%
11GPT-5.4 Nano (Reasoning, Low)100.0%$0.00062.4s100%
12Gemini 3.1 Flash Lite100.0%$0.00071.7s100%
13GPT-6 Luna100.0%$0.00024.1s100%
14GPT-5.4 Nano100.0%$0.00062.7s100%
15Gemma 3 4B100.0%$0.00015.4s100%
16Claude Haiku 5.5 (Reasoning, Medium)100.0%$0.00054.0s100%
17GPT-5.4 Nano (Reasoning)100.0%$0.00073.1s100%
18DeepSeek V4 Flash100.0%$0.00016.3s100%
19Claude Haiku 5.5 (Reasoning, Low)100.0%$0.00054.4s100%
20Gemma 3 12B100.0%$0.00016.6s100%
21Gemini 2.5 Flash100.0%$0.00121.7s100%
22Z.AI GLM 5.3 Flash (Reasoning, Low)100.0%$0.00017.1s100%
23GPT-6 Luna (Reasoning, Medium)100.0%$0.00036.7s100%
24GPT-6 Luna (Reasoning, High)100.0%$0.00047.0s100%
25GPT-4o Mini (temp=1)100.0%$0.00037.3s100%
26GPT-4.1 Mini100.0%$0.00084.9s100%
27Claude Haiku 5.5 (Adaptive)100.0%$0.00075.3s100%
28Qwen 2.5 72B100.0%$0.00028.1s100%
29GPT-4o Mini (temp=0)100.0%$0.00037.7s100%
30Mistral Medium 3.1100.0%$0.00104.9s100%
31Gemini 3 Flash (Preview)100.0%$0.00152.7s100%
32Arcee AI: Trinity Mini100.0%$0.00028.7s100%
33Mistral Large 3100.0%$0.00095.8s100%
34Grok 4.20100.0%$0.00163.4s100%
35Qwen 3.5 Plus (2026-02-15)100.0%$0.00125.4s100%
36GPT-5.4 Mini100.0%$0.00211.9s100%
37Qwen 3 235B A22B Instruct 2507100.0%$0.000310.9s100%
38Gemini 2.5 Flash Lite (Reasoning)100.0%$0.00117.2s100%
39Llama 3.1 70B100.0%$0.000412.3s100%
40Gemma 3 27B100.0%$0.000114.0s100%
41Mistral Large 4.0100.0%$0.00119.3s100%
42ByteDance Seed 1.6 Flash100.0%$0.000612.0s100%
43Claude Haiku 4.5100.0%$0.00272.2s100%
44GPT-5.6 Luna100.0%$0.00282.6s100%
45DeepSeek-V2 Chat100.0%$0.000613.3s100%
46DeepSeek V3 (2024-12-26)100.0%$0.000613.5s100%
47Xiaomi MIMO v2.6 Pro100.0%$0.001112.5s100%
48Laguna S 2.1100.0%$0.000118.3s100%
49Gemma 4 26B100.0%$0.000219.0s100%
50Mistral Medium 3.5100.0%$0.00372.2s100%
51Gemini 3.6 Flash (Reasoning, Minimal)100.0%$0.00381.9s100%
52Mistral Medium 3.5 (Reasoning)100.0%$0.00372.5s100%
53Hermes 3 405B100.0%$0.000916.2s100%
54DeepSeek V3 (2025-03-24)100.0%$0.000518.8s100%
55DeepSeek V4.1 Flash (Reasoning, High)100.0%$0.001614.2s100%
56Mistral Large 2100.0%$0.00345.8s100%
57GPT-4.1100.0%$0.00413.4s100%
58Gemini 3.5 Flash (Reasoning, Minimal)100.0%$0.00442.2s100%
59GPT-5.4 Mini (Reasoning)100.0%$0.00375.7s100%
60Thinking Machines Inkling100.0%$0.002013.9s100%
61Gemma 4 31B100.0%$0.000222.7s100%
62GPT-5.6 Luna (Reasoning, Medium)100.0%$0.00414.5s100%
63Muse Glimmer 30B (Reasoning, Medium)100.0%$0.002213.7s100%
64Qwen 3.7 Flash (Reasoning)100.0%$0.000622.4s100%
65DeepSeek V4 Flash (Reasoning)100.0%$0.000523.7s100%
66GPT-4o, Aug. 6th (temp=1)100.0%$0.00512.2s100%
67GPT-4o, Aug. 6th (temp=0)100.0%$0.00512.3s100%
68Writer: Palmyra X5100.0%$0.002714.3s100%
69GPT-6 Sol100.0%$0.00495.2s100%
70Gemini 4 Argon (Reasoning, Medium)100.0%$0.00523.9s100%
71Z.AI GLM 4.5 Air100.0%$0.001223.3s100%
72DeepSeek V4 Pro100.0%$0.001423.8s100%
73GPT-6.1 Sol (Reasoning, Medium)100.0%$0.00497.9s100%
74Muse Spark 1.1 (Reasoning, Minimal)100.0%$0.00546.5s100%
75Xiaomi MIMO v2.5100.0%$0.004015.1s100%
76Claude Sonnet 5.5 (Reasoning)100.0%$0.00653.3s100%
77WizardLM 2 8x22b100.0%$0.000633.4s100%
78Gemini 2.5 Flash (Reasoning)100.0%$0.004515.2s100%
79GPT-5.6 Terra100.0%$0.00712.7s100%
80Claude Sonnet 5100.0%$0.00656.6s100%
81GPT-5.4100.0%$0.00714.5s100%
82DeepSeek V3.2100.0%$0.000337.5s100%
83Nemotron 3 Super100.0%$0.000039.1s100%
84GPT-5.4 (Reasoning, Low)100.0%$0.00744.6s100%
85GPT-6 Sol (Reasoning, Medium)100.0%$0.00697.5s100%
86GPT-5.2100.0%$0.00765.6s100%
87Claude Sonnet 4.6100.0%$0.00813.6s100%
88Claude Sonnet 4.5100.0%$0.00814.0s100%
89Gemini 3.7 Flash (Reasoning, Medium)100.0%$0.006910.0s100%
90Z.AI GLM 5.3 Flash (Reasoning, Max)100.0%$0.000540.9s100%
91Xiaomi MIMO v2.5 Pro100.0%$0.004721.0s100%
92Grok 4.20 (Reasoning)100.0%$0.005218.8s100%
93Claude Sonnet 4100.0%$0.00814.9s100%
94Gemini 3.8 Flash (Reasoning, Medium)100.0%$0.00795.9s100%
95GPT-5.6 Terra (Reasoning, Medium)100.0%$0.00863.3s100%
96Z.AI GLM 5.2 (Reasoning, High)100.0%$0.005817.2s100%
97GPT-OSS 120B100.0%$0.001043.4s100%
98Muse Spark 1.1 (Reasoning, Medium)100.0%$0.00849.4s100%
99GPT-5 Mini100.0%$0.004230.5s100%
100Xiaomi MIMO v2.6 Flash100.0%$0.000450.1s100%
101MiniMax M3100.0%$0.002440.9s100%
102Qwen 3.6 35B100.0%$0.005028.5s100%
103MiniMax M2.5100.0%$0.001545.7s100%
104Qwen 3.8 Flash (Reasoning)100.0%$0.001248.5s100%
105Muse Spark 1.2 (Reasoning, Medium)100.0%$0.00987.8s100%
106Qwen 3.6 Flash100.0%$0.007221.7s100%
107Z.AI GLM 4.6100.0%$0.003838.3s100%
108Grok 4.5 (Reasoning, Low)100.0%$0.008119.3s100%
109ByteDance Seed 1.6100.0%$0.003839.9s100%
110Claude Sonnet 5 (Reasoning)100.0%$0.0109.7s100%
111Aion 3.5 (Reasoning, High)100.0%$0.006926.1s100%
112ByteDance Seed 2.0 Lite100.0%$0.003843.6s100%
113Claude Sonnet 5 (Reasoning, Low)100.0%$0.0119.7s100%
114Gemini 3 Flash (Preview, Reasoning)100.0%$0.009615.6s100%
115DeepSeek V4 Pro (Reasoning)100.0%$0.002052.4s100%
116Z.AI GLM 5 Turbo100.0%$0.008820.0s100%
117Aion 2.0100.0%$0.003645.3s100%
118Qwen 3.8 27B (Reasoning, XHigh)100.0%$0.006037.2s100%
119GPT-5 Nano100.0%$0.002454.7s100%
120Thinking Machines Inkling Small (Reasoning, High)100.0%$0.004644.0s100%
121GPT-5.1100.0%$0.01211.2s100%
122Muse Spark 1.3 (Reasoning, Medium)100.0%$0.01018.9s100%
123Grok 4.5 (Reasoning, High)100.0%$0.01116.1s100%
124o4 Mini100.0%$0.01117.5s100%
125Claude Opus 4.5100.0%$0.0144.7s100%
126Claude Opus 4.6100.0%$0.0144.9s100%
127GPT-5.6 Sol100.0%$0.0143.8s100%
128GPT-5.5100.0%$0.0144.0s100%
129Z.AI GLM 4.7 Flash100.0%$0.00141.1m100%
130Inception Mercury 296.4%$0.00152.1s74%
131Z.AI GLM 5.3 (Reasoning, Max)100.0%$0.009232.3s100%
132Cydonia 24B V4.196.4%$0.00039.0s74%
133MiniMax M2.7100.0%$0.004556.4s100%
134GPT-5.4 (Reasoning)100.0%$0.01413.0s100%
135Claude Opus 4.7 (Reasoning)100.0%$0.0163.3s100%
136Claude Opus 4.7100.0%$0.0163.5s100%
137GPT-5.4 Mini (Reasoning, Low)96.4%$0.00263.4s74%
138Grok 4.7 (Reasoning, High)100.0%$0.01324.2s100%
139ByteDance Seed 2.0 Mini100.0%$0.00141.3m100%
140Claude Opus 5100.0%$0.0166.9s100%
141Claude Opus 4.8 (Reasoning, Low)100.0%$0.0175.6s100%
142Claude Opus 4.8 (Reasoning)100.0%$0.0175.9s100%
143Mistral Small 4 (Reasoning)96.4%$0.001613.6s74%
144GPT-5.5 (Reasoning, Low)100.0%$0.0186.2s100%
145Kimi K3 (Reasoning, Low)100.0%$0.01425.8s100%
146Z.AI GLM 5100.0%$0.007756.2s100%
147Aion 3.0100.0%$0.01235.9s100%
148Z.AI GLM 4.7100.0%$0.00551.1m100%
149Aion 3.5 Mini (Reasoning, High)100.0%$0.00461.2m100%
150Grok 4.6 (Reasoning, High)100.0%$0.01430.7s100%
151Qwen 3.5 27B100.0%$0.01146.8s100%
152Aion 3.0 Mini100.0%$0.00461.3m100%
153Qwen 3.8 Max (Reasoning, XHigh)100.0%$0.01340.1s100%
154GPT-5.6 Sol (Reasoning, Medium)100.0%$0.0206.6s100%
155o4 Mini High100.0%$0.01626.9s100%
156Z.AI GLM 5.1100.0%$0.01154.4s100%
157Gemma 4 26B (Reasoning)100.0%$0.00121.7m100%
158Claude Opus 5 (Reasoning, Low)100.0%$0.0218.4s100%
159Laguna XS 2.192.9%$0.00029.1s65%
160Gemma 4 31B (Reasoning)100.0%$0.00091.8m100%
161GPT-5.5 (Reasoning, Medium)100.0%$0.0228.3s100%
162Claude Opus 5.5 (Reasoning)100.0%$0.0229.6s100%
163Claude Opus 4.6 (Reasoning)100.0%$0.0229.5s100%
164Gemini 3.6 Flash (Reasoning)100.0%$0.02210.6s100%
165Claude Opus 5 (Reasoning)100.0%$0.0239.4s100%
166Qwen 3.5 Plus (2026-04-20)100.0%$0.0111.1m100%
167Qwen 3.6 27B100.0%$0.01458.3s100%
168Qwen 3.5 122B100.0%$0.01841.8s100%
169Grok 4.3 (Reasoning)96.4%$0.006728.8s74%
170Mistral Large 4.0 (Reasoning)100.0%$0.0111.3m100%
171Gemini 2.5 Pro100.0%$0.02517.5s100%
172 Kimi K3 (Reasoning, High)100.0%$0.02137.1s100%
173Qwen 3.5 Flash96.4%$0.00301.0m74%
174Thinking Machines Inkling (Reasoning)100.0%$0.0121.4m100%
175Grok 4.392.9%$0.00163.5s48%
176Ember-1 (Reasoning, Max)100.0%$0.02526.7s100%
177Kimi K2.5100.0%$0.0111.6m100%
178Gemini 3.5 Flash (Reasoning)100.0%$0.02912.2s100%
179GPT-5100.0%$0.02541.3s100%
180Claude Sonnet 4.6 (Reasoning)100.0%$0.03120.5s100%
181Ministral 3B82.1%$0.00001.5s52%
182Qwen 3.5 397B A17B100.0%$0.00952.4m100%
183Qwen 3.7 Max100.0%$0.03051.4s100%
184Hy4 Preview (Reasoning, High)100.0%$0.0191.7m100%
185Muse Wonder (Reasoning, Medium)100.0%$0.0397.3s100%
186Gemini 3.1 Pro (Preview)100.0%$0.03431.5s100%
187Claude Opus 4100.0%$0.0416.4s100%
188DeepSeek V4 Flash 0731 (Reasoning, High)92.9%$0.00071.1m48%
189Gemini 3.5 Flash Lite (Reasoning)85.7%$0.00121.2s30%
190Gemini 3.5 Flash Lite (Reasoning, Minimal)85.7%$0.00121.2s30%
191Qwen 3.5 35B92.9%$0.01251.3s65%
192Kimi K2.6100.0%$0.0172.3m100%
193Qwen 3 32B85.7%$0.000525.5s30%
194Qwen 3.6 Max Preview100.0%$0.0301.6m100%
195Ministral 3 3B75.0%$0.00011.5s38%
196Qwen 3.5 9B92.9%$0.00131.9m48%
197DeepSeek V3.182.1%$0.000533.2s28%
198DeepSeek V4 Pro 0813 (Reasoning, High)92.9%$0.01359.4s48%
199Cohere Command R+ (Aug. 2024)64.3%$0.005414.3s15%
200Z.AI GLM 4.557.1%$0.001920.2s1%
201Nemotron 3 Nano67.9%$0.00172.2m19%
202Hermes 3 70B57.1%$0.00233.3m1%
98.41%

Individual Scenarios

Generic Prompt

▼

Expand all contractions

Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 Avg ▼
Claude Opus 5.5 (Reasoning)100100100100100100100100.0%
Qwen 3.8 Max (Reasoning, XHigh)100100100100100100100100.0%
GPT-6.1 Sol (Reasoning, Medium)100100100100100100100100.0%
Gemini 3.8 Flash (Reasoning, Medium)100100100100100100100100.0%
Ember-1 (Reasoning, Max)100100100100100100100100.0%
Muse Wonder (Reasoning, Medium)100100100100100100100100.0%
Z.AI GLM 5.3 (Reasoning, Max)100100100100100100100100.0%
GPT-5.6 Sol (Reasoning, Medium)100100100100100100100100.0%
Claude Opus 4.6 (Reasoning)100100100100100100100100.0%
Gemini 3.6 Flash (Reasoning)100100100100100100100100.0%
Gemini 3.7 Flash (Reasoning, Medium)100100100100100100100100.0%
Hy4 Preview (Reasoning, High)100100100100100100100100.0%
Muse Spark 1.2 (Reasoning, Medium)100100100100100100100100.0%
Qwen 3.7 Max100100100100100100100100.0%
DeepSeek V4.1 Flash (Reasoning, High)100100100100100100100100.0%
Z.AI GLM 5.3 Flash (Reasoning, Max)100100100100100100100100.0%
Qwen 3.8 Flash (Reasoning)100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100.0%
Muse Spark 1.3 (Reasoning, Medium)100100100100100100100100.0%
GPT-6 Luna (Reasoning, High)100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100.0%
Muse Spark 1.1 (Reasoning, Medium)100100100100100100100100.0%
Gemini 4 Argon (Reasoning, Medium)100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100.0%
Qwen 3.6 Max Preview100100100100100100100100.0%
GPT-5.5 (Reasoning, Medium)100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100.0%
Kimi K3 (Reasoning, High)100100100100100100100100.0%
Qwen 3.8 27B (Reasoning, XHigh)100100100100100100100100.0%
Claude Sonnet 5.5 (Reasoning)100100100100100100100100.0%
Grok 4.6 (Reasoning, High)100100100100100100100100.0%
GPT-6 Sol (Reasoning, Medium)100100100100100100100100.0%
Grok 4.7 (Reasoning, High)100100100100100100100100.0%
Kimi K2.6100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100.0%
GPT-5.6 Terra (Reasoning, Medium)100100100100100100100100.0%
GPT-6 Luna (Reasoning, Medium)100100100100100100100100.0%
Aion 3.5 Mini (Reasoning, High)100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100.0%
Mistral Large 4.0 (Reasoning)100100100100100100100100.0%
Kimi K3 (Reasoning, Low)100100100100100100100100.0%
Claude Opus 5 (Reasoning, Low)100100100100100100100100.0%
Aion 3.5 (Reasoning, High)100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100.0%
Xiaomi MIMO v2.6 Pro100100100100100100100100.0%
GPT-5100100100100100100100100.0%
Muse Spark 1.1 (Reasoning, Minimal)100100100100100100100100.0%
GPT-5 Mini100100100100100100100100.0%
Claude Opus 5 (Reasoning)100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100.0%
Kimi K2.5100100100100100100100100.0%
Thinking Machines Inkling (Reasoning)100100100100100100100100.0%
GPT-5.6 Sol100100100100100100100100.0%
GPT-5.1100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100.0%
GPT-5.6 Luna (Reasoning, Medium)100100100100100100100100.0%
MiniMax M3100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100.0%
DeepSeek V4 Pro 0813 (Reasoning, High)100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100.0%
GPT-6 Sol100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100.0%
Claude Opus 5100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100.0%
GPT-5.2100100100100100100100100.0%
Claude Haiku 5.5 (Adaptive)100100100100100100100100.0%
GPT-5.5100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100.0%
Muse Glimmer 30B (Reasoning, Medium)100100100100100100100100.0%
Aion 3.0100100100100100100100100.0%
o4 Mini High100100100100100100100100.0%
Z.AI GLM 5.3 Flash (Reasoning, Low)100100100100100100100100.0%
Qwen 3.7 Flash (Reasoning)100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100.0%
GPT-5.6 Terra100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100.0%
Thinking Machines Inkling Small (Reasoning, High)100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100.0%
Xiaomi MIMO v2.6 Flash100100100100100100100100.0%
Claude Opus 4100100100100100100100100.0%
GPT-4.1100100100100100100100100.0%
MiniMax M2.5100100100100100100100100.0%
Gemini 3.6 Flash (Reasoning, Minimal)100100100100100100100100.0%
Aion 2.0100100100100100100100100.0%
o4 Mini100100100100100100100100.0%
MiniMax M2.7100100100100100100100100.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100.0%
Mistral Medium 3.5 (Reasoning)100100100100100100100100.0%
Gemma 4 31B100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100.0%
GPT-5.6 Luna100100100100100100100100.0%
Gemma 4 26B100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100.0%
Claude Haiku 5.5 (Reasoning, Low)100100100100100100100100.0%
Mistral Large 4.0100100100100100100100100.0%
Mistral Medium 3.5100100100100100100100100.0%
GPT-5.4100100100100100100100100.0%
Mistral Large 3100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100.0%
Claude Haiku 5.5 (Reasoning, Medium)100100100100100100100100.0%
Gemini 2.5 Flash (Reasoning)100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100.0%
GPT-6 Luna100100100100100100100100.0%
Aion 3.0 Mini100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100.0%
DeepSeek V3.1100100100100100100100100.0%
DeepSeek V3.2100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100.0%
DeepSeek V4 Pro100100100100100100100100.0%
DeepSeek V4 Flash100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100.0%
Mistral Large 2100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100.0%
Grok 4.20100100100100100100100100.0%
Hermes 3 405B100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100.0%
GPT-5 Nano100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100.0%
Laguna S 2.1100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100.0%
Mistral Small 4 (Reasoning)100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100.0%
Qwen 3 235B A22B Instruct 2507100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100.0%
Gemma 3 12B100100100100100100100100.0%
Gemma 3 27B100100100100100100100100.0%
Laguna XS 2.1100100100100100100100100.0%
Mistral Small 4100100100100100100100100.0%
Thinking Machines Inkling100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100.0%
Cydonia 24B V4.1100100100100100100100100.0%
GPT-5.4 Nano100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100.0%
Ministral 3 14B100100100100100100100100.0%
Ministral 3 8B100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100.0%
Hermes 3 70B100100100100100100100100.0%
Arcee AI: Trinity Mini100100100100100100100100.0%
Gemma 3 4B100100100100100100100100.0%
Mistral NeMO100100100100100100100100.0%
Ministral 8B100100100100100100100100.0%
Grok 4.3 (Reasoning)1001001001001001005092.9%
Qwen 3.5 Flash1001001001001001005092.9%
Inception Mercury 21001001001001001005092.9%
DeepSeek V4 Flash 0731 (Reasoning, High)100100100100100100085.7%
Qwen 3.5 35B100100100100100505085.7%
Qwen 3.5 9B100100100100100100085.7%
Grok 4.3100100100100100100085.7%
Gemini 3.5 Flash Lite (Reasoning, Minimal)1001001001001000071.4%
Gemini 3.5 Flash Lite (Reasoning)1001001001001000071.4%
Qwen 3 32B1001001001001000071.4%
Nemotron 3 Nano100100100100500064.3%
Ministral 3B100100505050505064.3%
Ministral 3 3B5050505050505050.0%
Cohere Command R+ (Aug. 2024)5050505000028.6%
Z.AI GLM 4.510000000014.3%

Specific Prompt

▼

Expand all contractions

Model # 1 # 2 # 3 # 4 # 5 # 6 # 7 Avg ▼
Claude Opus 5.5 (Reasoning)100100100100100100100100.0%
Qwen 3.8 Max (Reasoning, XHigh)100100100100100100100100.0%
GPT-6.1 Sol (Reasoning, Medium)100100100100100100100100.0%
Gemini 3.8 Flash (Reasoning, Medium)100100100100100100100100.0%
Ember-1 (Reasoning, Max)100100100100100100100100.0%
Muse Wonder (Reasoning, Medium)100100100100100100100100.0%
Z.AI GLM 5.3 (Reasoning, Max)100100100100100100100100.0%
GPT-5.6 Sol (Reasoning, Medium)100100100100100100100100.0%
Claude Opus 4.6 (Reasoning)100100100100100100100100.0%
Gemini 3.6 Flash (Reasoning)100100100100100100100100.0%
Gemini 3.7 Flash (Reasoning, Medium)100100100100100100100100.0%
Hy4 Preview (Reasoning, High)100100100100100100100100.0%
Muse Spark 1.2 (Reasoning, Medium)100100100100100100100100.0%
Qwen 3.7 Max100100100100100100100100.0%
DeepSeek V4.1 Flash (Reasoning, High)100100100100100100100100.0%
Z.AI GLM 5.3 Flash (Reasoning, Max)100100100100100100100100.0%
Qwen 3.8 Flash (Reasoning)100100100100100100100100.0%
Grok 4.5 (Reasoning, High)100100100100100100100100.0%
Gemini 3.1 Pro (Preview)100100100100100100100100.0%
Muse Spark 1.3 (Reasoning, Medium)100100100100100100100100.0%
GPT-6 Luna (Reasoning, High)100100100100100100100100.0%
GPT-5.4 (Reasoning)100100100100100100100100.0%
Muse Spark 1.1 (Reasoning, Medium)100100100100100100100100.0%
Gemini 4 Argon (Reasoning, Medium)100100100100100100100100.0%
Z.AI GLM 5.1100100100100100100100100.0%
Qwen 3.6 Max Preview100100100100100100100100.0%
GPT-5.5 (Reasoning, Medium)100100100100100100100100.0%
Claude Sonnet 4.6 (Reasoning)100100100100100100100100.0%
Z.AI GLM 5.2 (Reasoning, High)100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning)100100100100100100100100.0%
Z.AI GLM 5 Turbo100100100100100100100100.0%
Kimi K3 (Reasoning, High)100100100100100100100100.0%
Qwen 3.8 27B (Reasoning, XHigh)100100100100100100100100.0%
Claude Sonnet 5.5 (Reasoning)100100100100100100100100.0%
Grok 4.6 (Reasoning, High)100100100100100100100100.0%
GPT-6 Sol (Reasoning, Medium)100100100100100100100100.0%
Grok 4.7 (Reasoning, High)100100100100100100100100.0%
Kimi K2.6100100100100100100100100.0%
Claude Opus 4.7 (Reasoning)100100100100100100100100.0%
GPT-5.5 (Reasoning, Low)100100100100100100100100.0%
GPT-5.6 Terra (Reasoning, Medium)100100100100100100100100.0%
GPT-6 Luna (Reasoning, Medium)100100100100100100100100.0%
Aion 3.5 Mini (Reasoning, High)100100100100100100100100.0%
Claude Opus 4.8 (Reasoning)100100100100100100100100.0%
Claude Opus 4.6100100100100100100100100.0%
Mistral Large 4.0 (Reasoning)100100100100100100100100.0%
Kimi K3 (Reasoning, Low)100100100100100100100100.0%
Claude Opus 5 (Reasoning, Low)100100100100100100100100.0%
Aion 3.5 (Reasoning, High)100100100100100100100100.0%
Claude Opus 4.8 (Reasoning, Low)100100100100100100100100.0%
Xiaomi MIMO v2.6 Pro100100100100100100100100.0%
GPT-5100100100100100100100100.0%
Muse Spark 1.1 (Reasoning, Minimal)100100100100100100100100.0%
GPT-5 Mini100100100100100100100100.0%
Claude Opus 5 (Reasoning)100100100100100100100100.0%
Qwen 3.5 397B A17B100100100100100100100100.0%
Grok 4.3 (Reasoning)100100100100100100100100.0%
Grok 4.5 (Reasoning, Low)100100100100100100100100.0%
GPT-5.4 (Reasoning, Low)100100100100100100100100.0%
Grok 4.20 (Reasoning)100100100100100100100100.0%
Kimi K2.5100100100100100100100100.0%
Thinking Machines Inkling (Reasoning)100100100100100100100100.0%
GPT-5.6 Sol100100100100100100100100.0%
GPT-5.1100100100100100100100100.0%
Claude Sonnet 4.6100100100100100100100100.0%
GPT-5.6 Luna (Reasoning, Medium)100100100100100100100100.0%
MiniMax M3100100100100100100100100.0%
Claude Sonnet 5 (Reasoning)100100100100100100100100.0%
Qwen 3.5 122B100100100100100100100100.0%
Claude Sonnet 5 (Reasoning, Low)100100100100100100100100.0%
GPT-6 Sol100100100100100100100100.0%
Qwen 3.5 27B100100100100100100100100.0%
Claude Opus 5100100100100100100100100.0%
Gemini 3 Flash (Preview, Reasoning)100100100100100100100100.0%
Claude Opus 4.7100100100100100100100100.0%
GPT-5.4 Mini (Reasoning)100100100100100100100100.0%
Qwen 3.5 Plus (2026-04-20)100100100100100100100100.0%
Gemma 4 31B (Reasoning)100100100100100100100100.0%
Claude Opus 4.5100100100100100100100100.0%
Z.AI GLM 5100100100100100100100100.0%
ByteDance Seed 1.6100100100100100100100100.0%
GPT-5.2100100100100100100100100.0%
Claude Haiku 5.5 (Adaptive)100100100100100100100100.0%
GPT-5.5100100100100100100100100.0%
Qwen 3.6 Flash100100100100100100100100.0%
DeepSeek V4 Pro (Reasoning)100100100100100100100100.0%
Gemma 4 26B (Reasoning)100100100100100100100100.0%
Muse Glimmer 30B (Reasoning, Medium)100100100100100100100100.0%
Aion 3.0100100100100100100100100.0%
o4 Mini High100100100100100100100100.0%
Z.AI GLM 5.3 Flash (Reasoning, Low)100100100100100100100100.0%
Qwen 3.7 Flash (Reasoning)100100100100100100100100.0%
Gemini 2.5 Pro100100100100100100100100.0%
Qwen 3.6 27B100100100100100100100100.0%
GPT-5.6 Terra100100100100100100100100.0%
DeepSeek V4 Flash (Reasoning)100100100100100100100100.0%
DeepSeek V4 Flash 0731 (Reasoning, High)100100100100100100100100.0%
Z.AI GLM 4.7100100100100100100100100.0%
Qwen 3.6 35B100100100100100100100100.0%
Z.AI GLM 4.6100100100100100100100100.0%
Claude Sonnet 4100100100100100100100100.0%
Claude Sonnet 4.5100100100100100100100100.0%
Thinking Machines Inkling Small (Reasoning, High)100100100100100100100100.0%
Claude Sonnet 5100100100100100100100100.0%
Xiaomi MIMO v2.6 Flash100100100100100100100100.0%
Claude Opus 4100100100100100100100100.0%
Qwen 3.5 35B100100100100100100100100.0%
GPT-4.1100100100100100100100100.0%
MiniMax M2.5100100100100100100100100.0%
Gemini 3.6 Flash (Reasoning, Minimal)100100100100100100100100.0%
Aion 2.0100100100100100100100100.0%
o4 Mini100100100100100100100100.0%
MiniMax M2.7100100100100100100100100.0%
Qwen 3.5 Plus (2026-02-15)100100100100100100100100.0%
Xiaomi MIMO v2.5 Pro100100100100100100100100.0%
Gemini 3.1 Flash Lite (Reasoning)100100100100100100100100.0%
Gemini 3.5 Flash (Reasoning, Minimal)100100100100100100100100.0%
ByteDance Seed 2.0 Mini100100100100100100100100.0%
Qwen 3.5 Flash100100100100100100100100.0%
Gemini 3 Flash (Preview)100100100100100100100100.0%
Gemini 3.1 Flash Lite (Preview)100100100100100100100100.0%
Mistral Medium 3.5 (Reasoning)100100100100100100100100.0%
Gemma 4 31B100100100100100100100100.0%
Gemini 3.1 Flash Lite100100100100100100100100.0%
GPT-5.6 Luna100100100100100100100100.0%
Z.AI GLM 4.5100100100100100100100100.0%
Gemma 4 26B100100100100100100100100.0%
GPT-OSS 120B100100100100100100100100.0%
Claude Haiku 5.5 (Reasoning, Low)100100100100100100100100.0%
Mistral Large 4.0100100100100100100100100.0%
Mistral Medium 3.5100100100100100100100100.0%
Gemini 3.5 Flash Lite (Reasoning, Minimal)100100100100100100100100.0%
Gemini 3.5 Flash Lite (Reasoning)100100100100100100100100.0%
GPT-5.4100100100100100100100100.0%
Mistral Large 3100100100100100100100100.0%
ByteDance Seed 2.0 Lite100100100100100100100100.0%
Claude Haiku 5.5 (Reasoning, Medium)100100100100100100100100.0%
Gemini 2.5 Flash (Reasoning)100100100100100100100100.0%
DeepSeek-V2 Chat100100100100100100100100.0%
Qwen 3.5 9B100100100100100100100100.0%
Xiaomi MIMO v2.5100100100100100100100100.0%
GPT-6 Luna100100100100100100100100.0%
Aion 3.0 Mini100100100100100100100100.0%
Claude Haiku 4.5100100100100100100100100.0%
Gemini 2.5 Flash Lite (Reasoning)100100100100100100100100.0%
DeepSeek V3 (2024-12-26)100100100100100100100100.0%
DeepSeek V3.2100100100100100100100100.0%
Z.AI GLM 4.7 Flash100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=0)100100100100100100100100.0%
DeepSeek V4 Pro100100100100100100100100.0%
DeepSeek V4 Flash100100100100100100100100.0%
Inception Mercury 2100100100100100100100100.0%
Nemotron 3 Super100100100100100100100100.0%
Mistral Large 2100100100100100100100100.0%
GPT-4.1 Mini100100100100100100100100.0%
GPT-4o, Aug. 6th (temp=1)100100100100100100100100.0%
Grok 4.20100100100100100100100100.0%
Hermes 3 405B100100100100100100100100.0%
Z.AI GLM 4.5 Air100100100100100100100100.0%
Gemini 2.5 Flash100100100100100100100100.0%
GPT-5.4 Mini100100100100100100100100.0%
GPT-5 Nano100100100100100100100100.0%
GPT-5.4 Nano (Reasoning)100100100100100100100100.0%
Laguna S 2.1100100100100100100100100.0%
DeepSeek V3 (2025-03-24)100100100100100100100100.0%
Gemini 2.5 Flash Lite100100100100100100100100.0%
Qwen 3 32B100100100100100100100100.0%
Writer: Palmyra X5100100100100100100100100.0%
Qwen 3 235B A22B Instruct 2507100100100100100100100100.0%
Grok 4.3100100100100100100100100.0%
GPT-4o Mini (temp=1)100100100100100100100100.0%
GPT-5.4 Nano (Reasoning, Low)100100100100100100100100.0%
Llama 3.1 70B100100100100100100100100.0%
Mistral Small 3.2 24B100100100100100100100100.0%
GPT-4o Mini (temp=0)100100100100100100100100.0%
Mistral Medium 3.1100100100100100100100100.0%
Gemma 3 12B100100100100100100100100.0%
Gemma 3 27B100100100100100100100100.0%
Mistral Small 4100100100100100100100100.0%
Thinking Machines Inkling100100100100100100100100.0%
Qwen 2.5 72B100100100100100100100100.0%
GPT-5.4 Nano100100100100100100100100.0%
WizardLM 2 8x22b100100100100100100100100.0%
ByteDance Seed 1.6 Flash100100100100100100100100.0%
Ministral 3 14B100100100100100100100100.0%
Ministral 3 8B100100100100100100100100.0%
GPT-4.1 Nano100100100100100100100100.0%
Arcee AI: Trinity Mini100100100100100100100100.0%
Cohere Command R+ (Aug. 2024)100100100100100100100100.0%
Gemma 3 4B100100100100100100100100.0%
Ministral 3 3B100100100100100100100100.0%
Mistral NeMO100100100100100100100100.0%
Ministral 8B100100100100100100100100.0%
Ministral 3B100100100100100100100100.0%
GPT-5.4 Mini (Reasoning, Low)1001001001001001005092.9%
Mistral Small 4 (Reasoning)1001001001001001005092.9%
Cydonia 24B V4.11001001001001001005092.9%
DeepSeek V4 Pro 0813 (Reasoning, High)100100100100100100085.7%
Laguna XS 2.1100100100100100505085.7%
Nemotron 3 Nano1001001001005050071.4%
DeepSeek V3.1100100100100500064.3%
Hermes 3 70B10000000014.3%