Text to Speech

There are over 200 voices. Mandarin Chinese has the most, in male, female and many styles, plus Sichuanese, Northeastern, Shaanxi, Beijing, Tianjin, Minnan and Cantonese voices. English, Japanese, Korean, French, German, Italian, Portuguese, Spanish and Russian each have dozens of voices, Vietnamese and Indonesian have a few, and more than 30 other languages have one voice each. Some voices speak several languages: pick a language first and the list shows only the voices that can speak it. You can preview every voice before choosing.

By the number of characters, at a price that depends on the voice (currently 5–50 credits per 1,000 characters). Before submitting you see roughly how many credits the text needs. Charge = characters ÷ 1000 × unit price, rounded up, at least 1 credit per run. You need to sign in; failed runs are refunded.

Up to 5,000 characters per run, about 20 minutes of speech. Split longer scripts into several runs.

Yes. Every run includes SRT subtitles aligned sentence by sentence, at no extra cost, ready for CapCut, Premiere and other editors.

Only convert text you have the right to use, and never use it to impersonate others, defraud anyone or for other illegal purposes. Voices marked Experimental are synthesized by an overseas service; availability and voices may change.