Vietnamese Auto Captions

Vietnam is one of the largest TikTok and Facebook markets in Southeast Asia, with podcasts, talk shows and a massive gaming and streaming scene producing long recordings that clippers cut daily. The diaspora in the United States, Australia and France follows along, and captions carry clips to viewers who scroll muted on a phone.

Vietnamese is written in the Latin alphabet with a dense system of diacritics: tone marks stacked on vowels that already carry their own marks, such as ấ, ở and ữ, plus the đ. Those marks change meaning entirely, so a transcript that drops them is unreadable to native viewers. Crayo transcribes the speech as Vietnamese when selected; northern and southern pronunciations are both transcribed, and names and English loanwords are the lines to check in the caption list.

A practical tip: stacked tone marks need vertical room and size, so use a caption style with generous line height, large text and two to four words per caption, and preview it on the clip before export because the fonts differ across the 32 styles and the stacked marks need a clean rendering at phone size.