Visual Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
49.8Accuracy
12
Feb 26, 2026
71.56Accuracy (Counting)
12
Feb 26, 2026
99.8Accuracy
12
Feb 26, 2026
266.2BD-RATE
11
Jun 15, 2026
68.52VQA Accuracy (Avg)
11
Jun 12, 2026
42LLM-judged Accuracy
11
Jun 11, 2026
58.7LLM-judged Accuracy
11
Jun 11, 2026
56.8LLM-judged Accuracy
11
Jun 11, 2026
48.4LLM-judged Accuracy
11
Jun 11, 2026
74.3LLM-judged Accuracy
11
Jun 11, 2026
70.5Accuracy
11
Jun 4, 2026
72.4Accuracy
11
Jun 4, 2026
43.05AP
11
May 22, 2026
42.43AP
11
May 22, 2026
93.2Accuracy
11
Jun 8, 2026
100Relative Performance
11
Apr 21, 2026
92Exact Match
11
Apr 21, 2026
88Exact Match
11
Apr 21, 2026
92.1CO1 Accuracy
11
Apr 6, 2026
99.6IS1 Accuracy
11
Apr 6, 2026
98.9Accuracy (VS1)
11
Apr 6, 2026
40.9Accuracy
11
Mar 25, 2026
66.42VQA-v2 Accuracy
11
Mar 20, 2026
72.8VQA BLEU (Single)
11
Feb 26, 2026
77.96Accuracy
11
Feb 26, 2026