Free Cantonese Speech-to-Text
Upload your Cantonese audio or video and get an accurate, editable transcript in minutes — automatic, browser-based and free to start.
Cantonese is a language in its own right, not a dialect of Mandarin — the two are not mutually intelligible. It is spoken by around 85 million people worldwide, centred on Hong Kong, Macau and Guangdong province, with large communities across Southeast Asia and in overseas Chinatowns. Everyday Hong Kong media, interviews and podcasts are recorded in Cantonese and written with traditional characters.
ConvertSpeech transcribes them automatically. Upload your audio or video and our speech-recognition engine writes out the spoken Cantonese in traditional Chinese characters with punctuation, ready to copy into a document, subtitle file or translation tool. It runs in your browser, needs no installation, and your first minutes are free with no account required.
What people transcribe Cantonese for
Journalists and interviews
Turn recorded Hong Kong interviews into quotable text in minutes instead of hours.
Researchers and students
Convert Cantonese lectures and oral-history interviews into searchable notes you can quote and analyse.
Creators and media
Generate Cantonese subtitles and show notes straight from your video or podcast audio.
Cantonese is not Mandarin
Cantonese and Mandarin share the same character-based writing system but are separate spoken languages with different vocabulary, grammar and sound systems — a Mandarin speaker cannot follow spoken Cantonese. Cantonese also has a much richer tone system, with six to nine tones depending on how you count, so getting the words right depends heavily on clear audio. Choose Cantonese, not Mandarin, when you upload a Cantonese recording.
Written Cantonese uses colloquial characters that never appear in formal Mandarin writing — 係, 唔, 嘅, 咗 and others — to capture how people actually speak. ConvertSpeech is set up for Cantonese and produces natural traditional-character text; as always, the largest factor in accuracy is recording quality — a clear signal with little background noise beats everything else.
Tips for accurate Cantonese transcription
- Record with a good microphone and minimal background noise for the best accuracy.
- Select Cantonese rather than Mandarin or auto-detect — they are different languages and the choice matters.
- One speaker at a time transcribes best; overlapping speech reduces accuracy.
- Casual Cantonese often mixes in English words, so review those spots in the transcript.
- For interviews, enable speaker detection (a free registered feature) to label each speaker.