Our new X account is live! Follow @wizwand_team for updates

E-commerce Performance Evaluation on In-House Ecom Dataset

96.82Shopping Guide Performance

GPT5-Thinking

Updated 4d ago

Evaluation Results

Method	Links
GPT5-Thinking 2025.12		96.82	69	91.37	70.33	74	81.33	89.67	78.67	77.89	70.25	73.47	58.33	84.27	85.67	79.33	69.67	46.8	76.29
CompassMax-V3-Thinking 2025.12		96.14	80.67	98.48	82.33	90.33	88	89	82.33	80.19	76.71	82.45	74.33	94.58	89	90.67	71.5	91.67	85.79
CompassMax-V3 2025.12		95.24	81.67	98.72	83	89.67	86.33	90.33	81.67	79.51	78.5	83.12	67	93.46	88.67	90	71.5	89	85.14
Deepseek R1 2025.12		91.01	71.38	98.8	72	78.33	80.78	84.33	82	79.25	68.83	72	60	90.72	82.67	81.94	71.05	79.55	79.1
gemini-2.5-pro 2025.12		87.9	77.33	97.36	72.67	77	82.67	89.33	82	73.81	68.9	81.82	61.67	89.46	81	90.33	68.17	93.67	80.89
Qwen3-235B-A22B 2025.12		87.69	67	87.86	69.67	82.67	75.67	88.67	79	74.49	66.67	76.43	61	90.65	85.67	84.67	68.17	84.67	78.27
GPT-4o 2025.12		87.26	67	90.65	75	76.33	81.33	81	82	75.51	67.11	64.31	62.67	70.24	76.67	83.67	66.67	61.67	74.65
GLM4.5 2025.12		84.62	71.33	94.48	73.33	80	78.67	86.67	80	71.77	65.55	74.07	58	84.52	83	80.67	65.5	69	76.54