AI Video

Turn a brief into a short video with AI images, an AI voice-over and captions, then use it in a post or keep it in your media library.

What is AI Video?

AI Video builds a short vertical (or square, portrait or wide) video from a few sentences. ScoutSocial writes a scene script, paints one still image per scene, records a voice-over, times captions to the speech and renders everything into an MP4 you can attach to a post or download.

There are no animated clips or avatars in this version. Each scene is a still image with a gentle camera move (zoom or pan) and a transition into the next scene.

How to open it

Create Posts: open the + menu next to the composer and choose Create AI video.
Media Assets: click the clapperboard button next to the upload button.

The feature needs the platform's OpenAI configuration. If the button is missing, the workspace's plan does not include it or the platform has not enabled it yet.

The five steps

1. Brief

Describe the video in a few sentences, or switch to I have the narration and paste the exact words the voice should say. Pick a target length (10 seconds to 3 minutes), the format and, optionally, a language. Without a language the script follows the language of your brief.

Write script costs a few AI Credits (the exact upper bound is shown under the form). It is the only step billed before you see the plan.

2. Scenes

Each scene has three fields:

Narration: what the voice says. Captions show the same words.
Image description: what the still image shows. Text, logos and real people are not allowed in AI images.
On-screen text: up to five words shown over the scene. Optional.
Image: by default an AI still is painted from the description. Click Use my image to pick a photo or graphic from your media library instead; that scene costs no image credits. Use an AI image switches back.

Reorder, add or remove scenes. Your plan sets the maximum number of scenes per video. Rewrite script writes a fresh script from the brief and replaces your edits.

3. Voice and look

Voice: 13 AI voices.
Image quality: Draft, Standard or Best. The number next to each level is the AI Credit price per image.
Motion and Transition: camera move per scene and how scenes connect.
Captions: on by default, with a style, a position and an optional highlight color for the word being spoken.
On-screen text: shown when a scene has on-screen text. Pick one of the video editor's text styles (for example the black box, the yellow marker or neon), a size (Small, Medium, Large) and a position. Auto keeps the text clear of the captions; Top and Bottom pin it. The spot the captions use is not offered, so the two never overlap. Changing the look and producing again costs nothing, because images, voice and captions are reused.

4. Produce

The estimate lists what the remaining steps will cost. Production takes a few minutes; you can close the dialog and keep working. A small card in the bottom-right corner shows the progress on every page and reopens the dialog when the video is ready.

5. Result

Preview the video, download it, or click Use in post to attach it to the current draft (offered when you started from Create Post). The video is also saved to Media Assets. New video starts the next one in the same dialog. Edit video takes you back to the scenes and the look of this video; producing again only bills the steps you changed.

To change a video later, open its latest version in Media Assets and click the Edit AI video button (clapperboard icon) in its detail view. Earlier versions stay in the library as ordinary videos.

Costs

AI Video is billed per step, at the real provider price, only when a step completes:

Step	Price
Script	At token cost, a few AI Credits
Images	Per image, at the chosen quality (same prices as Generate AI image)
Voice-over	Per started 100 characters of narration
Captions	6 AI Credits per minute of speech
Render	Free

A step that fails is not charged. If production stops, Try again reuses everything that was already produced, so you only pay for what was missing.

Disclosure

Videos made with AI Video are AI-generated media. When you publish a post with such a video, ScoutSocial sets the AI-content flag on YouTube and TikTok automatically. Meta platforms label AI content on their side.

Tips

Keep each narration to one or two spoken sentences. Long scenes make the image feel static.
Describe the image concretely: subject, setting, light, mood. "A steaming cup on a wooden desk at sunrise" beats "coffee".
If the safety system declines a scene's image, the error names the scene. Rewrite its description and start again; the other scenes are kept.