Understanding Support Metric in scikit-learn

I am testing a dataframe with the following balance:

And when I estimated on this data set using two different models I got those results:

Result 1:

Result 2:

Looking the documentation of Support, it states that:

"The support is the number of occurrences of each class in y_true."

So, should not support be 873 and 337 (the true ys)? Why both model testing on the same dataset returns different supports?