How do I make a video with an AI presenter?

In chat with the AI Social Media Manager, choose Create Video from the + menu. Generate with AI walks you through five steps — a presenter (or your own photo), a voice (or your own, cloned), the script, images and stock footage, then the editor for captions, text and music. Or Upload your own video and just add captions and music.

Open a chat with the AI Social Media Manager, click + and choose Create Video.

Two ways in

• Generate with AI — a video with an AI presenter and voice, from a script you write or have written. The five steps below.
• Upload your own — bring a video you already have (or one from your library) and go straight to the editor for captions, music and simple trims. Skip to the editor.

A short walkthrough of the editor appears the first time you reach it.

1 · The presenter

Pick your avatar from the gallery, or use Your custom avatars to upload a photo — a talking presenter is made from it for this video. For a custom one you can ask for background matting (the presenter isolated from the photo's background), choose an avatar style and layout, and set the motion intensity, with a motion prompt if you want to describe the movement.

2 · The voice

Search the voice list, filter by gender, language and accent, and play a Voice Preview before choosing. Or Clone your voice: upload a short audio sample, name it, and it appears under your custom voices for every video after this one.

3 · The script

• Write the exact words the presenter will say, or
• Generate with AI from a brief. Choose a length — ~15, ~30 or ~60 seconds — and the script is written to fit: about two and a half spoken words a second, so it never runs long.
• Then choose the video format — Reel (vertical), Landscape or Square — and the target platform. Under Quality settings: the rendering engine (Avatar IV is standard; Avatar V is more natural, and premium), the output resolution (720p, 1080p or 4K), how expressive the presenter is, and an optional motion prompt describing gestures.

4 · Images and scenes

Add what the video cuts to between presenter shots. Images: Upload from your computer, pick from My Library, or AI Generate — describe one scene in plain language and it's made in about 10–20 seconds, with your script alongside for reference. Videos: search free stock footage from Pexels. Each becomes a scene on the timeline.

5 · The editor

Everything sits on a timeline you can zoom and trim. From the side panel:

• Clips — stock footage is auto-populated to match the script while you work; swap any clip, fit it with Cover or Contain, and set its volume.
• Captions — built automatically from the script in short lines; pick one of eleven styles: classic, minimalist, neon, bold, karaoke, fire, shadow, gradient, highlight, pill, cinema.
• Text overlays — start from a template or add your own; twelve fonts, bold and italic, colour and outline, opacity, rotation, an entry animation, motion, and exact start and duration. Apply style to all text keeps them consistent.
• Background music — off by default; twelve tracks from upbeat to ambient, with a volume control.

Review Video renders the whole thing; progress shows as it goes.

Post it

Caption & Tags — a caption is written from your script once the video finishes rendering; edit it, hashtags and all — then choose one or more platforms. The video is saved to the Media Library and the post appears under Content to review, approve and publish like any other — see Review, approve and publish posts.

Drafts

Close before you're done and you're asked whether to save this video as a draft. Recent Drafts on the first step lists them; open one to resume, or delete it.