创建一个欢快的卡通风格 3D 动画短片场景。一只可爱的动物,毛发如雪豹,眼睛大而富有表现力,圆润可爱,在奇妙的冬季森林中快乐地奔跑。场景应包含圆形的积雪树、轻柔的雪花和透过树枝的温暖阳光。该生物的弹跳动作和灿烂的笑容应该传达出纯粹的喜悦。通过明亮、欢快的色彩和俏皮的动画,营造积极、温暖的氛围。不妨添加一些微妙、奇妙的音效,营造欢乐的冬日氛围。
[[["易于理解","easyToUnderstand","thumb-up"],["解决了我的问题","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["很难理解","hardToUnderstand","thumb-down"],["信息或示例代码不正确","incorrectInformationOrSampleCode","thumb-down"],["没有我需要的信息/示例","missingTheInformationSamplesINeed","thumb-down"],["翻译问题","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["最后更新时间 (UTC):2025-08-29。"],[],[],null,["# Vertex AI video generation prompt guide\n\nThis guide provides examples of the types of videos that you can create using\nVeo and shows you how to modify specific parts of a prompt to produce different\nresults.\n\nPrompt guide overview\n---------------------\n\nVertex AI Veo is a text-to-video and image-to-video generation model.\nTo use Veo, you must provide a prompt which is a text description of what you\nwant your generative AI model to generate.\n\nSafety filters\n--------------\n\nVeo applies safety filters across Vertex AI to help ensure that\ngenerated videos and uploaded photos don't contain offensive content. For\nexample, prompts that violate [responsible AI\nguidelines](/vertex-ai/generative-ai/docs/learn/responsible-ai) are blocked.\n\nIf you suspect abuse of Veo or any generated output that contains inappropriate\nmaterial or inaccurate information, use the [Report suspected abuse on\nGoogle Cloud\nform](https://support.google.com/code/contact/cloud_platform_report).\n\nBasics for writing prompts\n--------------------------\n\nGood prompts are descriptive and clear. To get your generated video closer to\nwhat you want, start with identifying your core idea and then refine your idea\nby adding keywords and modifiers.\n\nThe following elements should be included in your prompt:\n\n1. **Subject**: The object, person, animal, or scenery that you want in your video.\n2. **Context**: The background or context in which the subject is placed.\n3. **Action**: What the subject is doing (for example, walking, running, or turning their head).\n4. **Style** : This can be general or very specific. Consider using specific film style keywords, such as *horror film* , *film noir* , or animated styles like *cartoon* style render.\n5. **Camera motion**: Optional: What the camera is doing, such as aerial view, eye-level, top-down shot, or low-angle shot.\n6. **Composition**: Optional: How the shot is framed, such as wide shot, close-up, or extreme close-up.\n7. **Ambiance**: Optional: How the color and light contribute to the scene, such as blue tones, night, or warm tones.\n\nExamples of prompts and generated output\n----------------------------------------\n\nThis section presents several prompts and how the level of detail provided in\neach prompt lets you get closer to what you want in your video.\n\n### Icicles\n\nThis video demonstrates how you can use each of the elements in your prompt.\n\n### Man on the phone\n\nThese videos demonstrate how you can revise your prompt with more specific\ndetails about the video that you want Veo to generate.\n\n### Snow leopard\n\nThis prompt demonstrates the output that Veo might generate.\n\n### Running snow leopard\n\nThis prompt has more detail and demonstrates generated output that might be\ncloser to what you want in your video.\n\nMore tips for writing prompts\n-----------------------------\n\nThe following tips help you write prompts that generate your videos:\n\n- **Use descriptive language**: Use adjectives and adverbs to paint a clear picture for Veo.\n- **Provide context**: If necessary, include background information to help your model understand what you want.\n- **Reference specific artistic styles**: If you have a particular aesthetic in mind, reference specific artistic styles or art movements.\n- **Utilize prompt engineering tools** : Consider exploring prompt engineering tools or resources to help you refine your prompts and achieve optimal results. For more information, see [Introduction to\n prompting](/vertex-ai/generative-ai/docs/learn/prompts/introduction-prompt-design).\n- **Enhance the facial details in your personal and group images** : Specify facial details as a focus of the photo like using the word *portrait* in the prompt.\n\nAdd more details to prompts\n---------------------------\n\nThese examples show you how to refine your prompts to generate your videos.\n\n### Subject description\n\nThis example shows you how to specify a subject description.\n\n### Context\n\nThis example shows you how to specify context.\n\n### Action\n\nThis example shows you how to specify action.\n\n### Style\n\nThis example shows you how to specify style.\n\n### Camera motion\n\nThis example shows you how to specify camera motion.\n\n### Composition\n\nThis example shows you how to specify composition.\n\n### Ambiance\n\nThis example shows you how to specify ambiance.\n\n### Audio\n\nAudio is supported by `veo-3.0-generate-001` in\n[Preview](/products#product-launch-stages).\n\nClearly specify if you want audio. We recommend that you use separate sentences\nin your prompt to describe the audio. The following are examples of audio\ndescribed as part of a prompt:\n\n- Sound effects:\n\n - The audio features water splashing in the background.\n - Add soft music in the background.\n- Speech:\n\n - The man in the red hat says, \"Where is the rabbit?\" Then the woman in the green dress next to him replies, \"There, in the woods.\"\n\nUse reference images to generate videos\n---------------------------------------\n\nYou can bring images to life by using the image-to-video capability that Veo has\nand use your existing assets or Imagen to generate something new.\n\nWhen using image-to-video, we recommend the following:\n\n- Ensure that actions and speech descriptions align with each subject present\n in the input image.\n\n- When multiple subjects are present in the input image, clearly specify which\n character is performing an action or speaking. To differentiate between\n multiple characters in an input image, use the most distinguishing\n descriptive details available. For example:\n\n - The man in the red hat.\n - The woman in the blue dress.\n\nNegative prompts\n----------------\n\nNegative prompts can be a powerful tool that helps specify what elements to keep\nout of the video. Describe what you want to discourage the model from generating\nby describing what you want the model to generate. Follow these tips:\n\n- ❌ Don't use instructive language or words like *no* or *don't*. For example,\n \"No walls\" or \"don't show walls\".\n\n- ✅ Do describe what you don't want to see. For example, \"wall, frame\", which\n means that you don't want a wall or a frame in the video.\n\nAspect ratios\n-------------\n\nVertex AI Veo video generation supports the following two aspect\nratios:\n\n### Widescreen - aspect ratio of 16:9\n\nThis is a prompt example of the widescreen with an aspect ratio of 16:9.\n\n### Portrait - aspect ratio of 9:16\n\nThis is a prompt example of portrait with an aspect ratio of 9:16."]]