Chinese Transcription

Transcribe Chinese audio to text, free

Upload a Chinese recording and get a clean, readable transcript in seconds. Output is a continuous character stream with punctuation, matching normal Chinese typography. Simplified characters are produced by default. Free to start, no account required.

New uploads are paused until it finishes, so the service stays fast for everyone. You can leave this page — the result will be waiting in your queue.

Open my queue →

MP3, M4A, WAV, FLAC, OGG, MP4 · up to 20 MB as a guest

Transcribing your audio…

How accurate is Chinese transcription? Tested

What to expect

Quick read-through

Most sentences come out right; skim for names and numbers.

We ran 20 Chinese recordings from Google's FLEURS dataset — native speakers reading everyday news-style sentences — through the same Whisper large-v3-turbo setup this site uses, with Chinese selected, on 30 October 2026, and compared every transcript with what was actually said.

A typical sentence from the test

Said: 桥下垂直净空 15 米。该项目于 2011 年 8 月完工,但直到 2017 年 3 月才开始通车。

Transcript: 桥下垂直近空15米该项目于2011年8月完工但直到2017年3月才开始通车

Clear read speech is an easy case. Interviews, phone calls, several people talking at once and music under speech produce more errors. FLEURS sentences © Google, CC BY 4.0. How we measured · All 40 languages compared

What makes Chinese hard to transcribe

Written Chinese has no spaces between words, so accuracy is measured in character error rate rather than word error rate. Mandarin also carries four tones plus a neutral tone, and homophone density is extreme — dozens of distinct characters can share one pronunciation, meaning the model has to pick from meaning, not sound.

What your Chinese transcript looks like

Output is a continuous character stream with punctuation, matching normal Chinese typography. Simplified characters are produced by default.

Timestamped output 中文
[00:00 → 00:04] 大家好,今天的会议现在开始。
[00:04 → 00:08] 首先我们来看上个月的数据。

An illustration of how Chinese transcripts are formatted — script, punctuation and timestamps. Not the output of any particular recording.

How accurate to expect

Mandarin from broadcast, lecture and podcast sources transcribes well. Accuracy falls on regional accents from outside the Beijing standard, and on recordings where speakers switch into a local dialect.

What usually needs fixing

  1. 1 Homophone selection — the sounds are right but the character chosen carries a different meaning, most visibly in names.
  2. 2 Company and brand names, where the correct characters are a business decision rather than a phonetic one.
  3. 3 Numbers and measure words running together, since Chinese counts with a classifier attached to the number.

Getting better results

  • → Say numbers and dates clearly — Chinese number words are short and easily merged with neighbouring characters.
  • → Recordings that mix Mandarin with English product names transcribe fine; you do not need to separate them.

Who transcribes Chinese

International firms transcribing supplier and factory calls for their China operations
Academic researchers working with Mandarin-language fieldwork recordings
Media companies subtitling Chinese-language interviews for global distribution

Questions about Chinese transcription

Transcripts come out in Simplified characters. Converting to Traditional afterwards is a lossless one-step change.
Cantonese has its own language token and is recognised separately from Mandarin, though with lower accuracy given less training data.
The transcript comes out Simplified. Converting to Traditional is a straightforward one-step change in any text editor that supports it.
Yes, using full-width Chinese punctuation rather than Latin marks, so the text reads as properly typeset Chinese.

Need the opposite — Chinese text to speech?

Turn a Chinese script into a voiceover with 14 Chinese AI voices — preview each one free in your browser.

Chinese voices →

Transcribe other languages

Other languages people transcribe most on SpeakScribe.

SpeakScribe transcribes 32 languages with dedicated guidance, and 99 in total via auto-detect. See all audio-to-text options →

Transcribe your Chinese audio now

Upload a file and get a timestamped transcript in seconds — free, no account needed.

Open the transcriber