Hear2Text LogoHear2Text

Chinese

中文

Transcribe Chinese Audio to Text

Upload Mandarin audio or video, or paste a public link, and get a Chinese transcript in characters with timestamps, ready to edit, search and export. Standard Mandarin works best. The free plan gives you 120 minutes every week.

Timestamps and speaker labels on every plan

See It In Action

Here is a short clip of a Mandarin meeting as Hear2Text transcribes it, with an English translation below.

Original Audio

"大家好,感谢参加今天的会议。我们先回顾一下项目进度。"

Transcription

"大家好,感谢参加今天的会议。我们先回顾一下项目进度。"

Translation: "Hello everyone, thank you for joining today's meeting. Let's start by reviewing the project's progress."

Characters, punctuation and timestamps

Transcripts are written in Chinese characters, typically Simplified, with full-width Chinese punctuation and no spaces between words, the way Chinese text is normally read. There is no Pinyin. Accuracy reaches up to 99% on clear audio; homophones, rare names and technical terms are where you will make most corrections. Every segment has a timestamp, so you can click it, listen again and fix the character in the browser.

Try Free Now

Built for Mandarin, honest about the rest

Hear2Text recognises Chinese automatically, so you do not pick a language before uploading. Standard Mandarin, whether from Beijing, Taipei or Singapore, is where results are strongest. A regional accent, fast speech or a noisy room lowers accuracy. Cantonese, Hokkien, Shanghainese and other Chinese languages are not reliably supported, and we would rather tell you that than let you find out on a two-hour recording.

Try Free Now

Translate Chinese transcripts into English

On the paid plans, a finished transcript can be translated into English or any of more than 60 languages. That helps teams who do not read Chinese follow a supplier call or a lecture. Recordings that mix Mandarin and English are handled in part: the language is detected once per recording, short English terms usually come through, and long English stretches may be transcribed poorly.

Try Free Now
EnglishEspañol

Mandarin, accents and other Chinese languages

There are no dialect-specific models: the same model hears every recording. Standard Mandarin works best, regional accents lower accuracy somewhat, and other Chinese languages are not reliably supported.

Mandarin (Putonghua)

Mainland China

Standard Mandarin is what the model handles best. Clear Putonghua from a close microphone, such as a news broadcast, lecture or interview, gives the most accurate transcripts.

Taiwanese Mandarin

Taiwan

Mandarin as spoken in Taiwan transcribes well on clear audio. Local vocabulary and names may need a correction, and the characters may not come out in the Traditional script you expect.

Cantonese

Hong Kong and Guangdong

Not something we promise. Cantonese is a separate spoken language, and recordings in it often come back as Mandarin-style text or with many errors, so test a short clip before relying on it.

Singaporean Mandarin

Singapore

Clear Singaporean Mandarin usually transcribes well. Frequent switches into English, Malay or Hokkien words can be dropped or misheard, so check those passages.

Shanghainese

Shanghai

Shanghainese (Wu) is not a supported variety and usually transcribes poorly. Mandarin spoken with a Shanghai accent works much better than the dialect itself.

Hokkien

Fujian and Taiwan

Not supported in any reliable way. Speech in Hokkien (Taiwanese) is usually transcribed poorly; Mandarin passages in the same recording fare much better.

How to transcribe Chinese audio

Upload a Mandarin recording or paste a public link, check the characters against the audio, then export or translate the transcript.

Step 01

Add your Mandarin recording

Upload an audio or video file up to 500 MB and 6 hours, record in the browser or the app, or paste a public YouTube or Vimeo link.

Step 02

Chinese is detected automatically

You do not choose a language. Hear2Text recognises the Mandarin and writes it out in Chinese characters with a timestamp on every segment.

Step 03

Check names and homophones

Read through the characters, click a timestamp to replay any line, and fix wrong characters in names or technical terms right in the browser.

Step 04

Export or translate

Download TXT on the free plan. Paid plans add SRT, VTT, Word and PDF, plus translation of the transcript into English and more than 60 other languages.

Who transcribes Chinese with Hear2Text

Anyone who needs Mandarin speech as searchable Chinese text: business teams, students, researchers, journalists and creators, whether they read Chinese or rely on the English translation.

Creators on Chinese video

Turn a Mandarin vlog, interview or livestream replay into Chinese text for captions, video descriptions and articles; SRT subtitle files are on the paid plans.

12.5K

Teams working with China

Record a Mandarin call with a supplier or partner, transcribe it afterwards, and translate the transcript into English so colleagues who do not read Chinese can follow the decisions.

Researchers and journalists

Transcribe Mandarin interviews and focus groups for coding and quoting, and use timestamps to go back to the original audio before you cite a sentence.

Mandarin learners

Transcribe a podcast episode or TV clip, read the characters while replaying each line from its timestamp, and compare with an English translation on the paid plans.

What you get with a Chinese transcript

Up to 99% on clear audio

Clear standard Mandarin from a close microphone gets the best results. Noise, crosstalk and strong regional accents lower accuracy.

Simplified characters

Output is typically in Simplified characters with no switch for Traditional. Edit or convert the text if your readers need Traditional script.

Context for homophones

Characters are chosen from the whole sentence, not from the syllable alone, so common homophones come out right. Check rare names and terms.

Speaker labels on every plan

Every plan labels who is speaking, as Speaker 1, Speaker 2 and so on, which makes Mandarin meetings and interviews easier to follow. On the paid plans you can rename them.

Characters, not Pinyin

Transcripts contain Chinese characters only. For Pinyin, export the text and use a separate converter.

Frequently Asked Questions

Does it work for Cantonese as well as Mandarin?
No, we only promise Mandarin. Standard Mandarin from mainland China, Taiwan and Singapore transcribes well on clear audio. Cantonese is a separate spoken language, and recordings in it are often transcribed badly or rendered as Mandarin-style text. If you have Cantonese audio, try a short clip on the free plan first and judge the result before transcribing anything long.
Can I choose between Simplified and Traditional characters?
No, there is no setting for it. Chinese transcripts typically come out in Simplified characters, and the script is not something you can switch before or after transcription. If you need Traditional characters for readers in Taiwan or Hong Kong, you can edit the text in the browser, or export it and convert it with a separate converter tool before publishing.
How does it deal with Chinese homophones?
Mostly well, through context. Mandarin has many syllables that sound alike, and the model chooses characters from the surrounding sentence rather than from the sound alone. Common words come out right; rare names, company names and technical terms are where a wrong character slips in. Read those passages, click the timestamp to hear the audio again, and correct the character in the editor.
Can I get the transcript in Pinyin?
No. Transcripts are written in Chinese characters only, with no Pinyin and no tone marks. If you are learning Mandarin and want Pinyin, copy or export the text and run it through a separate Pinyin converter. Paid plans can also translate the transcript into English, which helps when you want to follow the meaning line by line alongside the characters.
Can I translate a Chinese transcript into English?
Yes, on the paid plans. Once the Mandarin transcript is ready, you can translate it into English or any of more than 60 languages from the transcript page. The free plan covers the Chinese transcript itself and TXT export. Machine translation makes a good first draft, but have anything you publish checked by a fluent reader, especially names, idioms and technical terms.
What audio and video formats can I upload?
Any common format works: MP3, WAV, M4A, AAC, FLAC, OGG, OPUS, WEBM, MP4, MOV, MKV, AVI and others, up to 500 MB and 6 hours per recording. You can also paste a public link from YouTube, Vimeo, TikTok, Instagram or a direct media file. Videos behind a login cannot be fetched, so download those yourself and upload the file.

Turn your next Mandarin recording into text

Upload a Mandarin recording or paste a public link, and read an editable, timestamped Chinese transcript in minutes, with 120 free minutes every week.

Get Started Free