Catt

Record and transcribe on iPhone and Android. Same account.

Get
.mp3 中文

Convert MP3 to text in Chinese (Mandarin)

Drop your MP3 file. Get a Chinese (Mandarin) (中文) transcript with speaker labels, timestamps, and an AI summary written in Chinese (Mandarin), not translated from English.

Free account • No credit card • Files up to 20 GB

A real Chinese (Mandarin) recording, transcribed

We uploaded a 46-second test recording to the live site on November 19, 2026, with the language set to Chinese (Mandarin). Against the text that was read aloud, the transcript had 2.5 errors per 100 characters. These are the lines it returned, unedited:

  1. 00:00Speaker 1报告警告称,没有人能保证目前在伊拉克采取的任何行动能够阻止宗派战争不断增长的暴力或走向混乱。
  2. 00:20Speaker 2因为远离大陆,哺乳动物无法长途跋涉而来,使得巨龟成为克隆群岛主主要的食草动物。
  3. 00:28Speaker 3亚马逊河也是地球上最宽的河流,有些河段的宽度可达六英里。
  4. 00:38Speaker 1迦南没有大森林,所以木材非常昂贵。
What was read aloud (the reference text)

报告警告称,没有人能保证目前在伊拉克采取的任何行动能够阻止宗派战争、不断增长的暴力或走向混乱。

因为远离大陆,哺乳动物无法长途跋涉而来,使得巨龟成为科隆群岛主要的食草动物。

亚马逊河也是地球上最宽的河流,有些河段的宽度可达6英里。

迦南没有大森林,所以木材非常昂贵。

Errors are words (characters for Chinese, Japanese and Thai) swapped, missed or added, after removing punctuation and case. The recording and reference text are from FLEURS by Google, released under the Creative Commons Attribution 4.0 licence; the transcript is ours. Long dashes are shown as short dashes.

The same recording as a MP3 file

We saved that Chinese (Mandarin) recording as MP3 with common settings to show what you would upload.

Length
00:45
File size
718 KB
Bitrate
128 kbps
Audio
MP3, 16 kHz, mono
No account (100 MB)
about 109 minutes
Any account (20 GB)
about 372 hours

The same 00:45 in every format we tested

  • OGG: 189 KB
  • M4A: 325 KB
  • AAC: 326 KB
  • MKV: 380 KB
  • MP4: 397 KB
  • MOV: 397 KB
  • WebM: 642 KB
  • MP3: 718 KB
  • WMA: 747 KB
  • FLAC: 1.7 MB
  • AVI: 2.2 MB
  • FLV: 3.4 MB
  • WMV: 3.8 MB
  • WAV: 3.9 MB

Files made on November 21, 2026. Your own files can be larger or smaller: size depends on the bitrate, the sample rate and, for video, the picture.

MP3 files in Chinese (Mandarin): what we handle

Most MP3 files come from podcasts, voice memos, audiobooks, music files. We decode MP3 directly (no manual conversion to MP3 first) and transcribe it with the Chinese (Mandarin) setting, which covers speakers from China, Taiwan, Singapore.

Native 中文 script (with diacritics)
Speaker labels for multi-person audio
AI summary written in Chinese (Mandarin)
Action items, decisions, email draft
Loudness + silence handled
Click-to-seek transcript + inline edit

FAQ

How good is MP3 to Chinese (Mandarin) transcription?

On November 19, 2026 we uploaded a 46-second Chinese (Mandarin) test recording to the live site with the language set to Chinese (Mandarin). The transcript had 2.5 errors per 100 characters against the text that was read aloud. Covers Mandarin (Putonghua), written in Simplified or Traditional characters. You can fix any word in the editor while you listen; check names and numbers before you rely on a transcript.

Is converting MP3 to Chinese (Mandarin) text really free?

Yes. A free account transcribes files with no daily limit, each up to 30 minutes, and your first file can run up to an hour. Speaker labels, reading, copying, the full editor and downloads in every format are free, and a free account makes 10 AI notes a month (action items, Chinese (Mandarin) email drafts, AI summaries). The Weekly Pass and paid plans add them to every file. Full-length files from $8.33/mo billed yearly.

What MP3 files can I transcribe?

Any MP3 file. Without an account you can upload up to 100 MB, about 109 minutes of MP3 at the 128 kbps of our test file. Any free account takes files up to 20 GB.

Will the output be in Chinese (Mandarin) script?

Yes. Our AI outputs Chinese (Mandarin) in its native script (中文), with proper diacritics and punctuation. The AI summary is also written in Chinese (Mandarin), not translated from English.