Skip to content

Chinese Voice Typing

0 characters · 0 characters

Autosaved in your browser — nothing is uploaded or stored on our servers. Voice recognition is handled by your browser and never stored by us.

WorldTyping’s free Chinese voice typing lets you speak Mandarin and get Simplified Chinese characters in your browser using your device’s own speech recognition — no installation, no account, and no audio ever reaching our servers.

Simplified only, and Mandarin only

Two hard limits worth knowing before you start. Recognition is requested as zh-CN, so it returns Simplified characters — there is no setting that produces Traditional, and writing for Taiwan or Hong Kong means converting afterwards.

And it is Mandarin. Cantonese is not a variant here but a different spoken language, with its own vocabulary and grammar; speaking it will not give you slightly worse results, it will give you the wrong language.

It skips pinyin entirely

Almost everyone types Chinese through pinyin: enter the romanisation, then choose the character from a candidate list. That second step is the part that takes the time.

Dictation does not use pinyin at all. It goes from sound to characters, choosing them from context, and there is no candidate list to work through. For fluent speakers it is much faster than typing — and it means the errors arrive already committed rather than as a wrong selection you can see coming.

Tone, and then homophones anyway

Mandarin distinguishes words by tone, so an unclear tone is a different word before character selection even begins. And even with the tone right, a syllable maps to many characters: 是, 事, 市, 试 all read shì.

Context usually resolves it. When it does not, what you get is a well-formed sentence containing the wrong word — no typo to spot, nothing visually wrong. Proofread dictated Chinese for sense; a spelling-style scan will find nothing because there is nothing of that kind to find.

Where the audio goes

The microphone is your browser’s. On Chrome and Edge the audio is sent to the browser vendor’s speech service and text comes back; WorldTyping is never in that path and never receives audio. The text stays in your browser, saved locally so a reload does not lose your draft.

That is also why dictation needs an internet connection, and why a browser without Mandarin speech support will not respond to the button.

Frequently asked questions

Can I get Traditional characters?

Not from dictation here. Recognition is requested as zh-CN and returns Simplified characters. There is no setting that produces Traditional, so if you are writing for Taiwan or Hong Kong you will be converting afterwards rather than dictating directly.

Does it understand Cantonese?

No. zh-CN is Mandarin. Cantonese is a different spoken language with its own vocabulary and grammar, and speaking it here will not produce sensible text — not merely a lower accuracy rate, but the wrong language.

Why did it write the wrong character?

Tone and homophones. Mandarin distinguishes words by tone, and even with tone correct there are many characters per syllable — shì alone covers 是, 事, 市, 试 and more. The recogniser picks from context and is usually right; when it is wrong you get a grammatical sentence with the wrong word, so read for meaning rather than scanning for typos.

Is my voice recorded or sent to WorldTyping?

No. Recognition is handled by your browser, and on Chrome and Edge the audio goes to the browser vendor’s speech service. WorldTyping never receives audio — only the text that comes back, which stays in your browser.

Which browsers support Chinese dictation?

Chrome and Edge on desktop and Android, with an internet connection. Mandarin is well supported there. Safari on iOS and macOS varies by version. If the microphone button does nothing, the browser is declining rather than the page failing.

Can I paste the Chinese text into Word or WeChat?

Yes. It is standard Unicode, so it pastes into Word, Google Docs, Gmail, WeChat, WhatsApp or a web form with no font to install.

Related tools

Found a bug or missing a feature? Send feedback — it goes straight to the maker.