# Cantonese Text to Speech > Kenerate AI's Cantonese text to speech reads written Cantonese aloud in 9 native Cantonese voices on MiniMax Speech 2.8 HD, from Professional Host (F) and (M) to Narrator and Kind Woman. Paste the text, pick a Cantonese voice, set the language to Cantonese, and download an MP3. Each clip takes up to 10,000 characters. URL: https://kenerateai.com/cantonese-text-to-speech Publisher: Kenerate AI (https://kenerateai.com) Last updated: 29 September 2026 ## Key facts - Native Cantonese voices: 9 on MiniMax Speech 2.8 HD - Voices: Professional Host (F), Professional Host (M), Narrator, Wisel Professor, Gentle Lady, Kind Woman, Cute Girl, Playful Man, Indifferent Staff - Language setting: Cantonese (Language & output) - Characters: Traditional or simplified - Per clip: Up to 10,000 characters - Output: MP3, WAV or FLAC, no watermark - Demo "Professional Host (F)" (Cantonese, MiniMax Speech 2.8 HD): script: 各位觀眾,大家好!歡迎收睇今晚嘅節目。 (Good evening, everyone! Welcome to tonight's programme.) - Demo "Narrator" (Cantonese, MiniMax Speech 2.8 HD): script: 好耐好耐以前,喺一條細細嘅漁村入面,住咗一個好鍾意唱歌嘅漁夫。 (A long, long time ago, in a small fishing village, there lived a fisherman who loved to sing.) - Demo "Kind Woman" (Cantonese, MiniMax Speech 2.8 HD): script: 多謝晒你今日嚟幫襯,我哋下個禮拜再見! (Thank you so much for shopping with us today. See you next week!) - Demo "Playful Man" (Cantonese, MiniMax Speech 2.8 HD): script: 喂,今晚去邊度食飯呀?我好想食火鍋喎! (Hey, where are we eating tonight? I really want hotpot!) - Demo "Wisel Professor" (Cantonese, MiniMax Speech 2.8 HD): script: 今日我哋會講吓香港嘅歷史,同埋維多利亞港點樣變成今日嘅樣。 (Today we'll talk about Hong Kong's history, and how Victoria Harbour became what it is today.) - Demo "Professional Host (M)" (Cantonese, MiniMax Speech 2.8 HD): script: 列車即將到站,請小心月台空隙。多謝合作。 (The train is arriving. Please mind the gap between the train and the platform. Thank you.) ## Steps 1. Paste your Cantonese text — Up to 10,000 characters per clip on MiniMax and ElevenLabs Multilingual v2 (5,000 on Eleven v3 and Qwen3). 2. Pick a Cantonese voice — Filter the MiniMax library by Cantonese to see the 9 native voices, and set Language to Cantonese in Language & output. 3. Set the delivery — Emotion, speed and pitch on MiniMax; stability and audio tags on ElevenLabs; a style prompt on Qwen3. 4. Generate and download — MP3 (or WAV/FLAC on MiniMax), no watermark, saved to your history. ## In depth ### Cantonese is its own spoken language Cantonese is the Yue variety of Chinese spoken in Hong Kong, Macau and Guangdong, with tens of millions of speakers [1]. It shares most characters with Mandarin but not the pronunciation, and spoken Cantonese has its own grammar words — 係 for “is”, 唔 for “not”, 嘅 for the possessive — written with characters Mandarin rarely uses [2]. ### Six tones, and why the voice has to be native Cantonese is usually described with six contrasting tones (nine counting the checked tones), against Mandarin's four [1][3]. The same syllable with a different tone is a different word, so a voice trained on Mandarin can't fake it. That's why this page uses only native Cantonese voices. ### Which voice for which job A quick map of the demos: - TV, events and announcements — Professional Host (F), Professional Host (M). - Stories and audiobooks — Narrator. - Lessons and explainers — Wisel Professor. - Shops and customer messages — Kind Woman, Gentle Lady. - Casual vlogs — Playful Man. References: [1] Wikipedia — Cantonese: https://en.wikipedia.org/wiki/Cantonese [2] Wikipedia — Written Cantonese: https://en.wikipedia.org/wiki/Written_Cantonese [3] Wikipedia — Cantonese phonology: https://en.wikipedia.org/wiki/Cantonese_phonology ## FAQ Q: Is there a Cantonese text to speech voice? A: Yes. MiniMax Speech 2.8 HD has 9 native Cantonese voices in Kenerate's Voice Studio: Professional Host (F), Professional Host (M), Narrator, Wisel Professor, Gentle Lady, Kind Woman, Cute Girl, Playful Man and Indifferent Staff. Q: How is Cantonese different from Chinese text to speech? A: Both read Chinese characters, but a Cantonese voice speaks Cantonese (粵語) and a Mandarin voice speaks Mandarin (普通話). Pick a Cantonese voice and set Language to Cantonese; the Chinese text to speech page covers Mandarin and dialects. Q: Does it read traditional and simplified characters? A: Yes. Our demos use traditional characters, as Hong Kong does; simplified text is read too. Colloquial Cantonese characters such as 嘅、咗 and 喺 work best for natural speech. Q: Is Cantonese text to speech free? A: You can start free: new accounts get sign-up credits that never expire, and clips download without a watermark. MiniMax Speech 2.8 HD costs 30 credits per 1,000 characters; each demo line here cost 1 credit. Q: Can it translate English into Cantonese? A: No, it reads the Cantonese text you give it. Translate first, then paste the Cantonese. Q: Which Cantonese voice should I use? A: Professional Host (F) or (M) for announcements and presenting, Narrator for stories, Wisel Professor for lessons, Kind Woman or Gentle Lady for customer messages, and Playful Man for casual vlogs. Q: 粵語文字轉語音點樣用? A: 喺 Kenerate 嘅 Voice Studio 貼上廣東話文字,揀 MiniMax 九把廣東話聲其中一把,喺 Language & output 揀 Cantonese,然後生成同下載 MP3。新帳戶有免費 credits,冇水印,每段最多 10,000 字。 ## Related pages - Chinese Text to Speech: https://kenerateai.com/chinese-text-to-speech - Text to Speech: https://kenerateai.com/text-to-speech - Korean Text to Speech: https://kenerateai.com/korean-text-to-speech - Japanese Text to Speech: https://kenerateai.com/japanese-text-to-speech ## Sources - Kenerate Voice Studio (engines, voices and limits from the app's code): https://kenerateai.com/app/voice - MiniMax Speech 2.8 HD on WaveSpeed: https://wavespeed.ai/models/minimax/speech-2.8-hd - ElevenLabs Eleven v3 on WaveSpeed: https://wavespeed.ai/models/elevenlabs/eleven-v3 - ElevenLabs — text to speech documentation: https://elevenlabs.io/docs/overview/capabilities/text-to-speech - Wikipedia — Cantonese: https://en.wikipedia.org/wiki/Cantonese - Wikipedia — Written Cantonese: https://en.wikipedia.org/wiki/Written_Cantonese - Wikipedia — Cantonese phonology: https://en.wikipedia.org/wiki/Cantonese_phonology