A generated video starts as a silent clip. You can now ask for sound when you generate it, for one product or for a whole catalogue run: a voice reading a script, an instrumental music track, or (in beta) a talking avatar whose lips follow the speech. Sound is part of the same generation, so it does not cost anything extra.
The three kinds of sound
| Option | What you get | What it needs |
|---|---|---|
| Voiceover | A voice reads a script over the video. | A voice (one of the preset voices, or a custom voice you made in AI Studio) and a script of up to 1,000 characters. Leave the script empty and the video’s own script is read. |
| Music | An instrumental track made for the video, faded out at the end. | A short description of the music, for example “warm acoustic guitar, unhurried”. A catalogue run uses one track for every video in the run. |
| Lip-sync (beta) | A talking avatar: the avatar’s clip is re-rendered so its mouth follows the speech. | A model (avatar) in the video, plus a voice and script as for a voiceover. |
When the speech runs longer than the clip, the video loops to fit it, up to 60 seconds. The original audio of the clip is replaced by the sound you chose.
Generate a video with sound
- Set up the video as usual. In AI Studio, pick the product (or the products, for a catalogue run), describe the scene, and choose video.
- Choose the sound. Pick voiceover, music or lip-sync, then the voice and script or the music description. Choose no sound to keep the video silent.
- Generate and wait. The silent video renders first, then the sound is made and mixed in. A video with sound takes a little longer to finish than a silent one.
- Review. Play the finished video to check the sound, then add it to a collection the same way as any other generation.
If the sound does not arrive
- My video finished without sound. Did I lose the generation?
- No. If the voice, music or lip-sync step fails, you still get the silent video you paid for, and it is marked as having no sound with the reason. Generate again with sound if you want it.
- Does sound cost extra?
- No. The video is the one generation that counts. Making the voice or the music and mixing them in add no charge.
What counts against your allowance
Every plan includes a monthly allowance of image generations and a separate one for video, and it resets at the start of each calendar month. Product images, scenes, videos, custom models (avatars) and generated video covers all count, as one image or one video each. Music tracks and voice previews on their own do not. A generation that fails is given back.
Credits, once the allowance is used
When a month’s allowance for images or videos is used up, AI Studio can keep going on credits you have bought. Credits are only spent after the allowance, so the allowance always goes first, and your credit balance is reported alongside the allowance.
- Credits are bought on the same billing you use for your plan: as a one-time charge approved in your Shopify admin if you pay Idukki through Shopify, a one-off card payment otherwise, or on an invoice from our team if we bill you directly.
- Credits are added only after the payment is confirmed with Shopify or the card processor, never on the click alone.
- Our team can also add credits to your account, for example to settle an invoice.