Guides › Transcription
How to transcribe Mandarin Chinese audio to text
Research interviews, lectures, meetings, a Chinese video you want to subtitle: typing out Mandarin by hand takes hours, and not every transcription tool handles Chinese well. Here is how to get a Chinese transcript and timed subtitles from a recording, what affects accuracy, and what to check before you use the text.
Get a Chinese transcript and SRT subtitles: try Respeak (transcription is free) →
What you get
- A transcript in Chinese characters, split into sentences at the pauses.
- TXT: one sentence per line, for notes, quotes, a script or a translator.
- SRT subtitles: timed to each sentence, with long sentences split, ready for video editors and video platforms.
Step by step with Respeak
- Upload the audio or video: MP3, M4A, WAV and more, or an MP4 or MOV video. For a video, the page reads only its sound in your browser and uploads that, so a large file sends just a few megabytes.
- Transcribe. The recording is turned into Chinese sentences. Respeak also handles English, Japanese, Korean, Spanish, French, German and Portuguese.
- Fix what was misheard: double-click a sentence to correct a word. Names, places, jargon and numbers are the usual suspects. This changes the text only, not the sound.
- Export: click Export transcript and choose Transcript (TXT) or Subtitles (SRT).
Try it on your Mandarin recording →
Free, and limits: transcription is free up to 60 minutes a day, and exporting is free; you need an account. On the free trial a recording can be up to 10 minutes long, and up to 30 minutes after buying any credit pack. Split longer recordings into parts.
What affects accuracy
- Distance and room: clear speech close to the mic transcribes best. Echoey rooms and speakers far from the mic lower accuracy.
- Music and crosstalk: background music and several people talking at once make it worse.
- Accents and dialects: standard Mandarin works best. Of the Chinese languages Respeak supports Mandarin only, not Cantonese or others.
- Proper nouns: names of people, companies and products often come out with the right sound but the wrong characters. Check every one.
If you also need English
Respeak transcribes; it doesn't translate. Export the TXT (one sentence per line) and paste it into a translation tool. For English subtitles, translate the SRT line by line and keep its numbers and timings as they are, so the English lines still match the video.
More you can do with the same recording
In the same editor you can remove background noise (free), cut ums and long pauses from the transcript (free), and fix a sentence you got wrong in your own voice. See How to edit audio by editing text.
Common questions
Does it work with videos?
Yes. Upload the MP4 or MOV as it is: the page reads only its sound and uploads that. Other video formats up to 30 MB can be uploaded directly too.
Does it translate into English?
No. Respeak transcribes but does not translate. Export the transcript and use a translation tool.
Does it support Cantonese?
No. Respeak supports Chinese (Mandarin), English, Japanese, Korean, Spanish, French, German and Portuguese; of the Chinese languages, Mandarin only.
Does it label who is speaking?
No. The transcript is split into sentences without speaker names, so it works best with one main speaker.
Is transcription free?
Yes, up to 60 minutes a day, and exporting the transcript and subtitles is free too; you need an account.
Have Mandarin audio to transcribe? Upload it, check the names, export TXT or SRT. Transcription is free up to 60 minutes a day.