Did you click on the listen icons on the right side, those that activate the Hungarian parser? I’m Hungarian and Google’s “Erdős” and “Szilárd” are basically indistinguishable from common speech versions, while “Csíkszentmihályi” has only one minor flaw, namely that it leaves a bit too much space between Csík and szent.
Huh, the English parser ignores special marks completely, leading to notably awful pronunciations of those. Except for Csíkszentmihályi, funnily.
Did you click on the listen icons on the right side, those that activate the Hungarian parser? I’m Hungarian and Google’s “Erdős” and “Szilárd” are basically indistinguishable from common speech versions, while “Csíkszentmihályi” has only one minor flaw, namely that it leaves a bit too much space between Csík and szent.