Cantonese interviews, transcribed properly
Cantonese (廣東話): Cantonese is supported as its own language (yue) — distinct from Mandarin — with traditional-script output. Colloquial Cantonese particles (啦, 喎, 咩) appear in the output. Accuracy trails Mandarin; broadcast speech does noticeably better than street conversation, so plan a review pass for anything critical.
Interviews in particular: Speaker labels separate interviewer from interviewee, and word timestamps let you cite the exact moment. Verbatim quotes for articles and papers, or a clean Q&A ready to code in NVivo or ATLAS.ti.
How it works
Record well
Record each voice clearly — an external mic or per-speaker tracks make diarization near-perfect.
Upload it
Drop the file into Whipscribe or paste a link. 30 minutes free daily, no signup to try.
Read & export
Speaker-labeled, timestamped Cantonese text in minutes. Export TXT, SRT, VTT, or DOCX.
Typical uses: Hong Kong sermons, radio archives, family-history tape.
Frequently asked
Can Whisper transcribe Cantonese interviews?
Yes. Cantonese is supported as its own language (yue) — distinct from Mandarin — with traditional-script output. Colloquial Cantonese particles (啦, 喎, 咩) appear in the output. Accuracy trails Mandarin; broadcast speech does noticeably better than street conversation, so plan a review pass for anything critical.
How are speakers handled in a interview?
Speaker labels separate interviewer from interviewee, and word timestamps let you cite the exact moment.
How much does it cost?
About 3.3¢ per audio minute (~$2 an hour), 30 minutes free every day, no signup to try.
Is my recording private?
Yes — private infrastructure, never sent to a third-party AI service, no training on uploads. Always get participants' consent before recording.