How to Generate Video From a Text Prompt in CapCut

CapCut can generate short video clips from a text prompt or animate a still image into motion, which gives you b-roll you do not have to film. The output is short and stylized, so it works best for transitions, backgrounds, and accents rather than as your main footage.

What you need

• CapCut desktop or mobile, signed in
• A clear visual idea you can describe in one or two sentences
• Optionally, a reference image to animate

Step 1: Open the AI video tool

From the toolbox, choose Text to video (or Image to video if you have a starting picture). You land on a prompt box with style and aspect-ratio options.

Step 2: Write a specific prompt

Describe the subject, the camera move, the lighting, and the mood. Vague prompts give generic results. Compare 'a forest' against 'a slow drone shot gliding over a misty pine forest at sunrise, golden light through the trees.' The second one gives the model something to work with.

Weak:   city street

Strong: rainy neon city street at night, reflections on wet
        asphalt, slow forward dolly, cinematic, shallow depth
        of field

Step 3: Pick a style and ratio

Choose a visual style such as cinematic, anime, or realistic, and set the aspect ratio to match your project so you do not have to crop later. Then click Generate and wait for the render.

Step 4: Regenerate or refine

AI generation is a slot machine. If the first result has melting edges or wrong motion, tweak the prompt and generate again rather than settling. Generate a couple of variations and keep the best one.

Step 5: Add to the timeline and blend

Send the clip you like to the timeline. Trim it short, add a quick fade, and color-match it to your real footage so the generated shot blends in instead of standing out.

Result: a three-second cinematic forest flyover, generated from a sentence, slots in as an opening shot you never had to film.