MortgageTax

Evaluating reading and understanding tax certificates as images

ModelsModels with a result.97
Top resultBest result on this benchmark.72.1%Claude Opus 5
Top-3 spreadPoints from first to third.1.8 pts
UpdatedDate the source changed the results.1 Sep 2026

Result and cost

96 of 97 results

Results

97 results
#ModelResultThe score from the source.CostUS dollars to run one task.TimeTime to run one task.
01Claude Opus 572.1% ±0.9$0.0748.0 s
02Claude Fable 5.170.8% ±0.9$0.17215 s
03Claude Opus 4.770.3% ±0.9$0.08416 s
04Claude Sonnet 570.0% ±0.9$0.0829 s
05Claude Opus 4.869.9% ±0.9$0.10621 s
06Gemini 3.1 Pro Preview69.4% ±0.9$0.02424 s
07Gemini 3 Pro69.1% ±0.9$0.04126 s
08Claude Fable 568.9% ±0.9$0.17516 s
08Gemini 2.5 Pro68.9% ±0.9$0.004824 s
10GPT-5.568.8% ±0.9$0.07128 s
11Gemini 3 Flash Preview68.7% ±0.9$0.007513 s
11Gemini 3.5 Flash-Lite68.7% ±0.9$0.00434.9 s
11Claude Opus 4.568.7% ±0.9$0.09211 s
11Claude 3.7 Sonnet68.7% ±1.8$0.0165.7 s
15Claude Opus 4.6 (Adaptive)68.5% ±0.9$0.06127 s
16MiniMax M368.4% ±0.9$0.008327 s
17GPT-5.468.3% ±0.9$0.10345 s
17Qwen3.6-27B68.3% ±0.9$0.01742 s
19Gemini 3.5 Flash68.1% ±0.9$0.01811 s
20Gemini 3.1 Flash Lite Preview68.0% ±0.9$0.00285.4 s
20Qwen3.6 Plus68.0% ±0.9$0.01378 s
22Claude Sonnet 4.667.7% ±0.9$0.04530 s
22Claude Opus 4.5 Thinking67.7% ±0.9$0.17828 s
24Gemini 3.6 Flash67.6% ±0.9$0.029.2 s
25Qwen3.5 Flash67.4% ±0.9$0.002237 s
26GPT-5.6 Terra67.3% ±0.9$0.0236.3 s
26GPT-5.6 Luna67.3% ±0.9$0.004823 s
26GPT-5.6 Sol67.3% ±0.9$0.07113 s
29Gemini 2.5 Pro Exp 03.2567.2% ±0.9$0.00368.9 s
30GPT-5.267.1% ±0.9$0.0571 s
31GPT-5 mini66.9% ±0.9$0.004425 s
31Claude 3.7 Sonnet Thinking66.9% ±1.8$0.02414 s
33Gemini 3.7 Flash66.7% ±0.9$0.0114.1 s
34Kimi K2.5 Thinking66.5% ±0.9$0.01580 s
35Kimi K366.3% ±0.9$0.03979 s
36Qwen3.7 Plus66.2% ±0.9$0.007661 s
37Muse Spark 1.166.1% ±0.9$0.02225 s
38GPT-4.165.9% ±0.9$0.00625.0 s
39Kimi K2.665.8% ±0.9$0.0292 min
40o365.7% ±0.9$0.0121 s
41GPT-4.1 mini65.5% ±0.9$0.00284.6 s
41GPT-565.5% ±0.9$0.02963 s
43Muse Spark 1.265.4% ±0.9$0.02645 s
44Gemini 3.8 Flash65.3% ±0.9$0.04715 s
45Qwen3.8-27B64.9% ±0.9$0.01280 s
46o4-mini64.8% ±0.9$0.01213 s
47Grok 4.664.2% ±1.0$0.0148.5 s
48Claude 3.5 Sonnet64.1% ±1.8$0.0154.2 s
49Qwen3.8 Max64.0% ±0.9$0.02747 s
49Inkling64.0% ±0.9$0.05669 s
49Claude Sonnet 4.5 Thinking64.0% ±1.0$0.05248 s
52GPT-5.4 mini63.5% ±0.9$0.0452 min
53Gemini 2.5 Flash (Preview)62.6% ±1.0$0.00127.3 s
54Claude Sonnet 462.5% ±1.0$0.0168.4 s
55Inkling-Small62.3% ±1.0$0.0122 min
56Claude Haiku 4.5 Thinking62.2% ±1.0$0.02130 s
57Magistral Small62.1% ±1.0$0.004511 s
58Gemini 2.5 Flash Preview Thinking61.9% ±1.0$0.001218 s
59Grok 4.561.8% ±1.0$0.0152 min
60GPT-5.161.4% ±0.9$0.02646 s
60Gemma 4 31B61.4% ±1.0$042 s
62Mistral Small 3.161.2% ±1.0$0.000498.3 s
63Claude Opus 4.161.1% ±1.0$0.08810 s
64Qwen3.5 Plus Thinking60.8% ±1.0$0.01162 s
65Gemini 2.0 Flash Exp60.4% ±1.9$0.000132.8 s
66Gemini 2.0 Flash 00159.7% ±1.9$0.000273.8 s
67MiMo-V2.559.3% ±1.0$0.001533 s
68GPT-5.4 nano59.1% ±1.0$0.002615 s
69Claude Opus 458.6% ±1.0$0.08814 s
70Llama4 Maverick Instruct Basic58.5% ±1.0$0.000752.7 s
71Gemini 2.5 Flash Preview 04.1758.4% ±1.0$0.0023.6 s
72Llama 4 Scout 17B 16E Instruct57.8% ±1.0$0.000572.4 s
75GPT-4o57.4% ±1.9$0.00716.0 s
76Gemini 1.5 Pro 00256.8% ±1.9$0.0024.2 s
77Claude Opus 4.1 Thinking56.1% ±1.0$0.13624 s
78Gemini 2.5 Flash Lite Preview55.8% ±1.0$0.000488.0 s
79GPT-4o mini54.5% ±1.9$0.000826.1 s
80GPT-5 nano53.6% ±1.0$0.002333 s
81GPT-4.1 nano52.8% ±1.0$0.0013.9 s
82Mistral Large 352.1% ±1.0$0.00329.3 s
83Magistral Medium51.9% ±1.0$0.0226 s
84Claude Sonnet 4 Thinking49.9% ±1.0$0.05246 s
85Grok 4.348.3% ±1.0$0.0127 min
86Grok 4.20 (Reasoning)45.4% ±1.0$0.02411 s
87Grok 444.5% ±1.0$0.08665 s
88Gemini 1.5 Flash 00242.8% ±1.8$0.000122.1 s
89Grok 4.1 Fast (Reasoning)42.6% ±1.0$0.002447 s
90Grok 4 Fast (Reasoning)42.1% ±0.9$0.004614 s
91Llama 3.2 90B Vision Instruct Turbo38.8% ±1.9$0.000585.1 s
92Mistral Medium36.4% ±0.8$0.00277.1 s
93Grok 4.1 Fast Non Reasoning34.3% ±0.9$0.0018.9 s
94Mistral Medium 3.528.9% ±0.8$0.02468 s
95Grok 4 Fast Non Reasoning25.3% ±0.9$0.00354.5 s
96Llama 3.2 11B Vision Instruct Turbo23.3% ±1.6$0.0000922.5 s
97Grok 2 Vision3.5% ±0.7$0.00380.0 s