From prompt to download: the image and video generation workflow

A generation starts with a prompt and ends with a result you can inspect. This guide follows the current image and video workflow, including what to prepare before submitting and where to find completed media.

1. Choose image or video

Open the image generator for a still image or the video generator for a moving scene. The homepage composer also lets you choose between the two.

Set one goal for the first take: a product close-up, a portrait, or a short camera move. A focused brief makes it easier to decide what to change next.

2. Add references where you need them

For an image, add reference pictures if you want to guide the subject or visual direction. For a video, use the Start and End image positions, or choose the Reference Video workflow when available.

Upload your media or select it from the picker, then confirm with Select. Check that the thumbnails appear in the intended positions before submitting. Start and End serve different roles, so check both if you use them.

Use media you own or have permission to upload. A reference helps describe the request; it does not guarantee exact lettering, identity, or product geometry in the output.

3. Choose the model, then the settings

Image models and video models offer different controls. Choose the model first, then use the aspect ratios, resolutions, and durations shown for that selection.

For video, the selector includes Google Veo 3.1 Lite, Fast, and Quality, alongside supported Omni models. For images, the selector includes Nano Banana models. Check the current selector rather than assuming every model accepts the same inputs.

4. Sign in and review credits

Sign in with Google before generating. Review the estimated credits beside the generate control and check your prompt and references once more.

Buying a plan or credit pack and submitting a generation are separate actions. You can review the available options on the pricing page. Yearly plans grant the included annual credits together after payment.

5. Follow the task on the Create page

After an accepted submission, the Create page shows the task progress. Your prompt and reference media should remain available for that request. Keep the task open while it runs, or return later to view the saved generation history.

If a task reports an error, review its status before submitting another request. Each new submission is a separate generation with its own credit estimate.

6. Preview, download, and find it again

Preview the completed result on the Create page, then use its download action to save the file. Your own generated images and videos also appear in Assets. Use the available loading controls to view older results.

For a product video, check the product outline, label, reflections, and camera motion. For an image, check small details and text at full size. Revise the prompt and start a new generation when you need another version.

Prepare a video prompt or create a reference image for your next shot.