文字转视频

AI 提示词
0/2000
10s
6s30s
积分消耗💎30

示例视频

Kinetic porcelain rings rotate in a calm gallery light.

AI 提示词:

示例视频 1
示例视频 2
示例视频 3

媒体报道

Multiple AI modelsChoose from the text-to-video models currently shown in the selector.
Prompt-first workflowStart from a written scene without requiring an image upload.
Preview before downloadReview the completed clip in the workspace first.
Creations stay organizedCompleted generations are saved to your workspace.
文生视频 AI

将文字转化为电影级视频 Shape the motion.

用简单的英语描述场景,选择模型,即可生成广告、社交媒体和推介视频——无需摄影团队。

  1. 1Write a promptSet the scene and action.
  2. 2Choose a modelReview current availability.
  3. 3Set the formatUse the exposed controls.
  4. 4Generate & previewReview before download.

Prompt patterns

提示词 → 示例输出

每行展示了书面提示与示例结果的配对方式 —— 与上方的生成器流程一致。

01

Example 1

提示词创意
提示示例 1
示例片段

完整提示

场景: 阳光照射的沙漠公路上,红蓝金属漆半挂卡车高速漂移变形为擎天柱机器人 风格: 史诗、技术、工程力量 特色: 超现实主义渐进式机甲变形,35mm镜头,动态推入配合旋转平移

02

Example 2

提示词创意
提示示例 2
示例片段

完整提示

场景:近距离拍摄手部切开闪亮的陶瓷苹果 风格:ASMR,微观细节 特点:清脆的声音,木制切菜板,陶瓷质感

03

Example 3

提示词创意
提示示例 3
示例片段

完整提示

大理石上旋转的无线耳机的产品主视觉图,影棚灯光,极简背景。

Practical controls

顶级文生视频模型,尽在一处

在 Sora、Veo、Grok、Kling、WAN、Runway 等模型之间自由切换,无需离开工作区。

Sora 2 model showcase
Sora 2
Veo 3 model showcase
Veo 3
Grok model showcase
Grok
Kling 3.0 model showcase
Kling 3.0
WAN 2.6 model showcase
WAN 2.6
Runway model showcase
Runway
PixVerse model showcase
PixVerse
Vidu model showcase
Vidu

Model

Compare the current engines by the kind of scene and motion you want.

Duration

Choose a clip length when the selected model exposes that option.

Aspect ratio

Match the available frame shape to the intended placement.

Resolution

Select from the output sizes supported by the active model.

Where it fits

From a written idea to a visual draft

Text-to-video works well when the scene exists as language before you have source footage.

Campaign concepts

Turn a creative direction into short visual options for review.

Storyboard moments

Explore a shot before committing to production or animation.

Product ideas

Prototype a setting, camera move, or reveal from a written brief.

Social experiments

Test concise concepts in formats available for the selected model.

Prompt craft

Write prompts the model can stage

A compact, coherent direction is easier to interpret than a long list of competing events.

  1. 1Lead with one primary actionMake the main subject and its most important movement unmistakable.
  2. 2Name the camera behaviorTry a locked shot, slow push-in, pan, tilt, or orbit when it matters.
  3. 3Add atmosphere with restraintLighting, weather, and mood should support the action instead of competing with it.
  4. 4Remove contradictory directionsKeep pace, framing, and movement compatible within a short clip.

Before you create

A transparent generation workflow

Text-to-video is iterative. Current options, processing time, and results depend on the selected model and traffic.

Generation time varies

Model, settings, and traffic affect how long a job takes.

Results vary by prompt

Preview the clip and revise the direction when needed.

Credits appear first

Review the displayed credit estimate before you create.

References need permission

If a model accepts reference media, only upload files you may use.

Review the policies that apply to uploads and outputs:Privacy policyTerms of service

Questions, answered

文本生成视频常见问题

Useful details for planning a prompt and reviewing a generation.

Keep creating

Move between generation, transformation, control, and enhancement workflows.

准备好从文本生成了吗?

使用此页面顶部的生成器 —— 输入提示语,几分钟内即可完成创作。