Multimodal Instruction Following
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
93.1Score
35
Apr 21, 2026
1,382ELO
15
Feb 26, 2026
8.86Score
12
May 8, 2026
77.84SciQA Score
9
Mar 31, 2026
92Score
9
Feb 26, 2026
645.2English Metric
7
Feb 26, 2026
58.8Conversation
4
Feb 26, 2026