Share your thoughts, 1 month free Claude Pro on usSee more

Document Understanding on GRAPH2EVAL-BENCH

59.16F1 Score

GPT-4o

Updated 1mo ago

Evaluation Results

Method	Links
GPT-4o 2025.10		59.16	48.73	76.23	3,560.92
GPT-4o 2025.10		57.66	48.74	78.54	1,631.82
Qwen2.5-VL-72B 2025.10		57.3	48.37	70.94	2,394.19
Qwen2.5-VL-72B 2025.10		56.73	47.11	69.99	3,855.65
Deepseek-V3 2025.10		54.97	46.35	79.84	3,462.88
Deepseek-V3 2025.10		53.76	45.18	83.51	1,710.65
Deepseek-V3.1 2025.10		52.76	43.29	78.16	1,777.98
Deepseek-V3.1 2025.10		51.41	42.53	78.75	3,435.12
Qwen2.5-VL-32B 2025.10		43.41	38.13	50.08	3,779.32
GPT-4.1 mini 2025.10		42.02	42.02	82.02	3,593.13
Qwen2.5-VL-32B 2025.10		36.77	33	48.11	2,275.01
GPT-4.1 mini 2025.10		34.96	30.68	69.72	4,679.13
Qwen2.5-VL-7B 2025.10		34.96	25.48	69.73	3,732.27
Qwen2.5-VL-7B 2025.10		20.93	19.39	54.27	2,455.17