4 seconds will be deducted from your balance for each generated sound, regardless of its actual duration.
Equal balance between AI creativity and prompt adherence