Start with words
Write a scene from scratch. Describe what happens and how the camera moves, then choose landscape 16:9 or portrait 9:16.
Veo 3.1
Turn a written scene or reference images into video. Choose Veo 3.1 Fast or Quality, then refine your prompt before generating.
Describe the subject, action, camera movement and light. Start with text, animate a still image, or guide a scene with reference images—all in the same workspace.
Write a scene from scratch. Describe what happens and how the camera moves, then choose landscape 16:9 or portrait 9:16.
Upload one image as the starting frame, or add a second image to guide the ending frame. Describe the motion between them.
Use up to three reference images with Veo 3.1 Fast. Reference-image mode currently uses the 16:9 format.
210 credits per generation
Explore different ideas with the Fast option. Supports text, first/last frames and reference-image inputs.
875 credits per generation
Choose the Quality option for your next version. Supports text and first/last-frame inputs.
Name the subject and setting, then add one clear action. Upload images if you want to guide the result.
Select Fast or Quality, the aspect ratio and image input mode. Check the credit cost shown before generating.
Sign in with enough credits, generate, then review the result. Adjust one detail at a time for your next version.
These are editable prompt ideas, not previews of generated videos. Each prompt is written in English.
A quiet coastal village at sunrise, pastel houses reflected in still turquoise water. A small wooden boat drifts slowly across the harbor as the camera gently tracks from left to right. Soft morning light, realistic textures, restrained cinematic colors. No text, captions, logos or watermark.
A minimalist glass perfume bottle on a pale stone pedestal. The camera makes a slow, smooth orbit around the bottle while soft daylight moves through sheer curtains in the background. Subtle reflections, elegant product photography, clean neutral colors. No text, captions, logos or watermark.
Close-up of an adult woman with chestnut hair standing beside a sunlit window. She slowly turns toward the camera and gives a relaxed smile while a light breeze moves a few strands of hair. Gentle camera push-in, natural skin texture, shallow depth of field. No text, captions, logos or watermark.
No. You can create from text alone. For image-guided generation, upload a starting frame, optional ending frame, or reference images.
First/last-frame mode accepts one or two images and works with Fast and Quality. Reference-image mode accepts up to three images and currently requires Fast and 16:9.
Fast uses 210 credits per generation; Quality uses 875. The composer shows the selected cost before you generate.
Results can vary. Be specific about the subject, action and camera movement, and change one part of the prompt at a time when refining your video.