The “Taiwanese Minnan Input Method” supports both phonetic input and voice input, allowing users to convert speech into text.