In general, increasing the number of decision tree models always helps. Each new tree will give diminishing returns, thus if the difference between nine and ten models is very small, it is very unlikely that an eleventh model will make a big difference. How quickly the benefits decrease depends on your dataset.
Typical forests sizes are often 10, 30, or 100 models. BigML runs 10 decision trees by default, however you can explore and find the right number for your data. BigML is capable of creating ensembles with up to 1000 models.