Voice and background music

A narrated video uses an AI voice or a voice file you recorded. You pick the voice when you make the video; then in the editor you can change the voice style, read sentences again, cut sentences and set the music — all in the Sound tab.

Pick a voice when making the video

Step 1. Open the voice list

In the start form, open Voice and press it to open the Choose a voice window.

Step 2. Filter and listen

• Search by name or style (e.g. "deep", "storytelling").
• Filter by Gender, Accent and Use (Storytelling, Education, Ads, Social media…).
• Press ▶ to listen before you pick. Press the star to keep a voice under Saved voices.

The "Choose a voice" window: the search box, the Gender / Accent / Use filters, the voice list with its ▶ listen buttons and star buttons

Step 3. Pick the language and speed (optional)

Under Advanced: Video language (AI writes the script in it, and the voices are listed for it) and Reading speed (slower or faster than the voice reads).

Use your own voice

Pick the From your voice file way to start and upload your voice file. If you have a .srt / .vtt caption file, add it — the words on screen are then exactly yours (names, terms). Without one, AI hears and writes down the words (currently 1 credit a minute). AI cuts the scenes on your own voice.

Change the voice in the editor

Open the Sound tab (which parts show depends on the project):

• Voice-over: One voice for the video (pictures change with the words, the voice never breaks between scenes) or A voice per scene (set in the scene cards on the Content tab).
• Voice style: Steady (news, how-tos), Balanced (most videos), Expressive (storytelling), or Fine-tune. The style applies to what is read again from now on; for the whole script, press Read the whole voice again in this style.
• Read one sentence again: open the sentence list and press Read again to change a sentence's words or how it is read. Compare the Current reading with the New reading, then press Use this reading. The price shows before it reads; if it fails, nothing is charged.
• Cut: take a sentence out of the voice, free; later scenes move up.
• Say it like this…: teach the voice how to say a word, e.g. "sangtao.ai ⇒ sang tao dot A I". Shared by every project of your account, captions still show the written form, up to 200 lines.

To switch the whole video to another voice, ask in the Assistant tab — it sends voice cards to listen to, then you press Use this voice.

Background music

• When making the video: Background music offers Picked for the content (AI picks from the music library), Choose music or No music.
• In the editor: the music field in the Sound tab opens the music library — search, filter by mood (Chill, Upbeat, Business, Emotional, Cinematic, Vietnamese…), listen, then pick.
• Your music: where the template allows it, drop an MP3, M4A or WAV file (up to 20 MB, 10 minutes) and tick I have the right to use this music.
• Volume: set the voice, music and sound effects apart; turn on Lower the music while the voice speaks, or Balance automatically to go back to the template's mix.

The Sound tab: Voice style with Steady / Balanced / Expressive, the "Read one sentence again" list, and Volume with the Voice / Music / Sound effects sliders

See also

• Make a video from a template
• Editing in the editor
• Exporting a video