No specific max token limit for the system prompt the max tokens setting controls the maximum tokens the model can generate per response default is 250..