✪ Developer Region: Light yellow indicates models developed in Europe; light blue, the United States; light green, Taiwan; light purple, China; light beige, other parts of Asia; and light orange, Africa.
✪ Percentage Scores: Scores above 50% are shown in black, while scores below 50% are shown in red.
✪ Starting in May 2026, a Model Evaluation Date column has been added for reference. Starting in September 2026, models are ranked by their aggregate score across the three foundational evaluation benchmark.
