The first video is easier when the workflow is small and clear. This guide walks through one complete path instead of every tool at once.
Givon AI keeps the core workflow in one editor: create a project, describe a scene, generate variants, assemble scenes, add voice or captions, and export a ready file.
For the first run, choose a short idea with two or three scenes. A product demo, a short Reels concept, or a visual mood test is enough to understand how projects, materials, and tokens work.
What you need
- A Givon AI account.
- A simple idea for a 2-3 scene vertical video.
Step by step
- 1
Create a project
Open Projects and create a new video project. The editor opens with an empty material area and prompt bar.
- 2
Describe the first scene
Choose a video mode and model, then describe the subject, movement, camera, style, duration, and frame format.
- 3
Generate several variants
Do not stop at the first result. Two or three variants usually reveal which prompt and model direction works.
- 4
Add more scenes
Use a strong frame as a reference or start frame so the next scene keeps the same subject or style.
- 5
Add voiceover and captions
Use voiceover mode for narration, upload your own audio, or create captions from transcription before export.
- 6
Review and export
Watch the full video for rhythm, text readability, and audio timing, then export the final file.
FAQ
How many scenes should a first video have?
Two or three scenes are enough. The goal is to complete the full workflow quickly.
Can I change the model after the first scene?
Yes. Running the same prompt through another model is a good way to compare movement, style, speed, and token cost.
Should I add voiceover immediately?
Usually no. Build the visual rhythm first, then add voice and captions before the final export.
