7 voices · Northern Mandarin · Standard Mandarin · Cantonese (Hong Kong) · Taiwan Mandarin · 48 kHz
Chinese text to speech
Seven Chinese voices in four varieties: Standard Mandarin for news and commentary, a northern storyteller, Taiwan Mandarin for easy-going vlogs, and a Hong Kong Cantonese radio host. Paste your script in simplified or traditional characters and hear it read the way a native speaker would.
Listen
Every Chinese voice
Guoqiang 国强
Male · elderly · Northern Mandarin
“从前啊，山脚下住着一个放牛娃。有一天，他在河边捡到一块会发光的石头，这事儿可就说来话长喽。”
Once upon a time, a young cowherd lived at the foot of the mountain. One day he found a glowing stone by the river, and that is quite a long story.
- storytelling
- grandfatherly
- warm
Haoran 浩然
Male · young adult · Standard Mandarin
“漂亮！这一波团战打得太精彩了！对面还想偷龙，没门儿！兄弟们，我们直接推上高地！”
Beautiful! What an amazing team fight! They still want to steal the dragon? No way! Come on, guys, let’s push straight up to their high ground!
- gaming
- commentary
- energetic
Jun 俊
Male · adult · Standard Mandarin
“各位观众，晚上好。今天是国庆长假的第一天，全国各大景区迎来出游高峰，交通部门提醒大家错峰出行。”
Good evening. Today is the first day of the National Day holiday. Scenic spots across the country are seeing peak crowds, and transport authorities remind everyone to travel off-peak.
- news
- broadcast
- authoritative
Kaman 嘉敏
Female · adult · Cantonese (Hong Kong)
“各位聽眾，早晨！今日天氣好好，陽光充足，不過下晝可能有驟雨，出門口記得帶定把遮啊。”
Good morning, listeners! It’s a lovely sunny day today, but there may be showers this afternoon, so remember to take an umbrella when you head out.
- radio
- cantonese
- friendly
Mei
Female · young adult
“你好，欢迎使用语音合成工作室。”
Hello, and welcome to the speech synthesis studio.
- chinese
- gentle
- conversational
Xiaoyu 晓雨
Female · young adult · Standard Mandarin
“家人们，今天这款保温杯真的太好用了！早上倒的热水，到晚上还是烫的，颜色也超级好看，赶紧拍起来！”
Family, this thermos is just so good! Hot water you pour in the morning is still hot at night, and the colours are gorgeous too. Get yours now!
- advertising
- livestream
- lively
Yating 雅婷
Female · young adult · Taiwan Mandarin
“哈囉大家好，今天帶你們去逛我最愛的夜市！這攤的鹽酥雞真的超好吃，每次來都要排好久喔。”
Hi everyone, today I’m taking you to my favourite night market! The popcorn chicken at this stall is so good; every time I come there’s a long queue.
- vlog
- gentle
- friendly
About Chinese in Voxmint
Chinese is the deepest part of the catalogue after English. Jun reads like an evening news anchor, Haoran calls a game at full volume, Xiaoyu sells like a livestream host, and Mei keeps things soft and conversational. Guoqiang brings a northern Mandarin grandfather’s storytelling voice, Yating speaks Taiwan Mandarin with the relaxed warmth of a travel vlogger, and Kaman reads Cantonese the way a Hong Kong morning show sounds.
You can write your script in simplified or traditional characters. Yating’s and Kaman’s samples above were written in traditional characters and read straight from them. For Cantonese, write the way people actually speak it, with colloquial characters such as 嘅, 咗 and 喺; a script written in formal Mandarin wording will come out sounding like a news bulletin read in Cantonese rather than everyday Hong Kong speech.
Most characters with more than one reading, like 行, 长 or 重, are read correctly from context. When one slips, rewording the phrase usually fixes it faster than anything else. Numbers, dates and amounts are safest written the way you want them spoken: 二零二六年 or 两千零二十六年, not a bare 2026, because a year and a quantity are read differently.
Punctuation shapes the rhythm. A Chinese comma gives a short breath, a full stop a proper pause, and an exclamation mark lifts the energy, which is exactly how Haoran’s sample gets its excitement. Tone errors are rare on common vocabulary but can appear on unusual names and rare characters; if a name matters, listen to a take before you publish it and re-render or add a short phrase around it if it sounds off.
For dialogue, cast two or three voices rather than one. Guoqiang and Mei make a natural grandparent-and-child pair, Haoran and Xiaoyu work well as co-hosts, and Jun can frame a segment as the announcer. If none of the ready-made voices fits, describe the voice you need in the studio, including the variety of Chinese it should speak, and keep the take you like best.
Design or clone a Chinese voice
Describe the voice you want, for example:
A calm Mandarin-speaking woman in her early thirties with a clear, warm mid-pitched voice, speaking at an unhurried pace like a documentary narrator.
Or clone one from a few seconds of your own recording.
Use cases
What people make with Chinese voices
Short video and livestream clips
Energetic voices like Xiaoyu and Haoran are made for Douyin-style product clips, game highlights and fast captions that need to grab attention in the first second.
News, briefings and explainers
Jun’s broadcast delivery suits daily briefings, company announcements and explainer videos where clarity and authority matter more than personality.
Children’s stories and audiobooks
Guoqiang tells a story like a grandfather at bedtime; pair him with Mei for dialogue and you have a warm, two-voice story without booking a studio.
Cantonese and Taiwan audiences
Kaman and Yating let you reach Hong Kong and Taiwan listeners in the variety they actually speak, instead of defaulting to mainland Standard Mandarin.
Questions
- Does it do Cantonese?
- Yes. Kaman is a Hong Kong Cantonese voice, and you can design more Cantonese voices by describing one in the studio. Write your script in colloquial Cantonese characters for the most natural result.
- Should I write in traditional or simplified characters?
- Either works. Use the script your audience reads; the voice follows the characters you give it, and the Taiwan and Hong Kong samples on this page were written in traditional characters.
- Can I mix English words into a Chinese script?
- Short English words and brand names inside a Mandarin sentence are common and usually read naturally. For long English passages, switch to an English voice for that part.
- Which other Chinese dialects can it speak?
- The model behind Voxmint supports nine Chinese dialects. Today the catalogue has ready-made voices for Cantonese and Taiwan Mandarin; for the others, describe the voice you want and listen to a few takes.
- How do I control pauses and emphasis?
- Use punctuation: commas for short breaths, full stops for longer pauses, and exclamation marks for energy. Splitting a long script into shorter sentences also gives a clearer, more natural read.
- Can the voices sound more emotional?
- Wording and punctuation carry most of the emotion: an exclamation mark lifts the energy and a gentle phrase softens it. For a particular style, such as a soft, breathy whisper, describe it when you design a new voice.