Vietnamese AI voiceover

The Vietnamese voices, with samples

Each sample reads the same line so you can compare them fairly.

Vietnamese scripts, written not converted

With Vietnamese in the Language field, the director composes narration and on-screen copy in Vietnamese from your brief's facts, and the vi-VN locale binds every automatic voice. Register follows your description, bạn for a friendly spot, quý khách when formality fits, and northern-standard vocabulary is the default the voices match. A brief in English works exactly as well as one in Vietnamese, since the director writes rather than translates. Quoted Vietnamese in the brief, a dish name or slogan, passes through with its diacritics untouched.

The full cast, tonal language included

All eight premium characters exist in vi-VN. Puck and Laomedeia carry promos and lively social spots; Charon and Erinome read instructions evenly, and tutorials are bound to them by the same code-level rule as every language; Umbriel and Sulafat suit eateries and testimonial warmth; Iapetus and Aoede handle corporate reads. Tones are the voices' native ground, so nothing special is needed from you there. Scene-level overrides work through the voice dialog, with the entire live catalogue searchable behind the automatic eight.

Six tones, short words, quick captions

Vietnamese sentences build from many short words, which makes the word-by-word caption particularly lively; the highlight moves fast, and a 48 character sentence split arrives sooner than its visual length suggests. Keep sentences compact and the pill reads beautifully. Scene timing behaves as everywhere: the scene stretches to its clip, dynamic pace floats an overrun across the next cut, delay in seconds and speed steps 0.85x to 1.25x sit per scene, and the soundtrack ducks itself to 22 percent while the voice speaks.

Diacritics in text, sounds in the voice

Always write full diacritics in any narration you edit by hand; stripped tone marks change words, and the voice reads what is written, so bán and bàn are different sentences. Foreign names respell into Vietnamese phonetics in the narration box, about half a credit per regenerated clip, and the respelling doubles as the caption text on that scene. Numbers spelled as words, một trăm nghìn đồng, avoid digit-phrasing surprises. The em dash never appears in generated narration by prompt rule; commas and full stops carry Vietnamese rhythm better anyway.

Captions in this script

Vietnamese words are short and spaced, so captions type on word by word in tight sync with the voice, timings taken from the clip. Tone marks are the font question: of the families the looks use, Inter, Plus Jakarta Sans, Montserrat, IBM Plex Sans, Lexend, Lora, Work Sans and Space Grotesk all ship the Vietnamese subset, so most looks render stacked diacritics like ệ and ở natively. The exceptions are Outfit, DM Sans and Poppins, used by the Aurora, Reverie, Linen, Cerulean, Blossom and Retro Pop looks, where Vietnamese glyphs fall back to a system font mid-caption. Pick from the larger group and every caption stays in the look's own font.

Questions people ask

Related pages