XiaoMin
Cantonese (China) Text to Speech and AI Voice Generator
Cantonese (China) text to speech for Guangzhou and the rest of Guangdong: type or paste a script in Simplified characters, then press Play.
Edit the text, pick a voice and press Play. Listen as often as you like before you sign up. Type in Simplified characters, and keep Cantonese words such as 嘅, 系 and 唔 where you want the line to sound spoken rather than read.
3 voices
Cantonese (China) text to speech voices, male and female
Listen2It has 3 Cantonese (China) voices, 1 male and 2 female. Press Use this voice to load one into the tool. Filter by gender.
Chi-Wai
Mei-Ling
Proof
Used by teams and publishers

“In the biggest languages, the voices sound nearly human.”
Review on Capterra , 1 December 2022
“…an impressive natural-like array of voices and great customer support.”
Review on Capterra , 27 September 2022
This page is for Cantonese as it is spoken in Guangdong, with Guangzhou as its centre. It sounds close to Hong Kong Cantonese, but the writing differs. Mainland readers learn Simplified characters, so write 广州 and 粤语 rather than 廣州 and 粵語. Choose this page for viewers in Guangzhou or Foshan, and the Hong Kong page when your subtitles will be in Traditional characters.
Guangdong scripts often switch register with the channel. Ads, food videos and short clips use spoken Cantonese, with words such as 乜嘢 for what and 咁 for so. Formal notices are usually written in standard Chinese and simply read aloud in Cantonese. Pick one register and keep to it through the whole script.
Local place names can trip a voice up, because their Cantonese readings differ from the usual ones. 番禺 is read pun1 jyu4, not with the common reading faan1 for 番, and the 埔 in 黄埔 is bou3. Play every district and street name in your script. Fix any slip once in the pronunciation dictionary.
Three steps
How to convert Cantonese (China) text to speech in three steps
Add your script
Type or paste Cantonese (China) text, or import a .txt, .docx or .pdf file up to 500KB, or a web page by URL.
Pick a voice and shape the delivery
Play the samples, choose a voice, add pauses where the listener needs a breath, and set how any word is read with the pronunciation dictionary.
Generate and download
Export an MP3, a voice and music mix, or SRT captions.
What's free
Free plan
Everything you need to hear your script
- The Free plan costs $0.
- Every voice is included.
- Preview voices as often as you like.
- A free account gets 2,500 credits once, not every month. That is about 2,500 words of text to speech.
- Download your audio as an MP3.
- The Free plan is for personal and evaluation use. Paid plans add commercial rights.
Answers
Frequently asked questions
Yes. You can play any voice on this page before you sign up. The full library has 1,411 voices across 160 languages and dialects.
Write or import your Cantonese (China) script, choose a voice and press Generate. To download the MP3, create a free account: downloads work on the Free plan. You can also export a voice and music mix or SRT captions.
Yes. The Free plan costs $0 and includes every voice, with a one-time allowance of 2,500 credits, about 2,500 words. It is for personal and evaluation use.
Yes. Cantonese (China) has 1 male and 2 female voices. Use the gender filter above the voice list and try a few with your own first line.
Yes, on a paid plan: paid plans include commercial rights, and the Free plan is for personal and evaluation use only. Starter costs $5 for the first month, then $19 a month.
Preview the line first. If a number, date or abbreviation comes out wrong, write it the way it should be spoken, or add it to the pronunciation dictionary, which is on every plan. One entry fixes that word across the whole project.
The two sound much alike, and speakers understand each other easily. The main differences are in writing and in some vocabulary. Guangdong uses Simplified characters and borrows more words from Mandarin, while Hong Kong uses Traditional characters and mixes in more English. Pick the page that matches where your listeners live and how your subtitles are written.
It depends on the channel. Food videos, adverts and social clips sound natural in spoken Cantonese, written with characters such as 咗 and 啲. Official notices and training are usually written in standard Chinese and read aloud in Cantonese. Preview your first line both ways and keep the one that fits.
The rest of the studio
More with Listen2It
Turn your Cantonese (China) script into a voice over
Paste your script into the tool above, pick a voice and hear it before you sign up.
Try it with your script