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

Overall Multi-benchmark Performance on AI2D, ChartQA, POPE, MME, and MMBench

100Average Performance

Qwen2.5-VL-7B

76.39282.52188.6594.779Jun 30, 2026
Updated 24d ago

Evaluation Results

MethodLinks
2026.06
100
2026.06
98.9
2026.06
98.5
2026.06
97.5
2026.06
94.1
2026.06
92.2
2026.06
91.5
2026.06
90
2026.06
88.2
2026.06
83.8
2026.06
82.7
2026.06
81.2
2026.06
77.3