Mandarin Text to Speech

Searches for cantonese vs mandarin usually come down to one thing: spoken, they are different languages, and a Mandarin speaker cannot follow Cantonese without learning it. Cantonese has six tones against Mandarin’s four and keeps final consonants that Mandarin lost, so even shared words sound unrelated. Written Simplified Chinese is largely common to both, which is why the confusion persists. Mandarin is the official spoken standard of mainland China and Taiwan; Cantonese dominates Hong Kong, Macau and Guangdong.

Only if the audience reads rather than listens. For written subtitles, Simplified and Traditional conversion handles most of it. For audio, a Mandarin voice over in Hong Kong is a foreign language rather than an accent, and it will be received that way. Cantonese voices are listed on the Chinese language page.

The language is the same and the details differ enough to notice. Taiwan writes in Traditional characters, uses some different vocabulary, and the accent softens certain retroflex sounds that mainland Mandarin keeps sharp. These voices are mainland zh-CN. For a Taiwanese audience the Chinese language page is the better starting point.

Characters are the ones to paste. Pinyin without tone marks is ambiguous by design, since one syllable maps to many characters, while characters carry both the tone and the word, so they give a cleaner read every time. Convert before generating and the tones come out right.

Chinese groups numbers in units of ten thousand rather than thousand, so a figure punctuated in Western style can be read in the wrong blocks. For anything where the number matters, write it in characters. Measure words are the other common slip, since the correct one depends on the noun.

Yes, on a paid plan. Audio generated on a paid plan is royalty free and cleared for advertising, client work, YouTube monetisation, e-learning and phone systems. Video exported on the free plan keeps a small Fliki watermark.