AI Video Generator

Start with a prompt, photos, or reference media. Shape the motion and keep every version in your works.

Credit quoteAfter sign-in
Sign in

Preview

Bring a travel photo to life

Seedance 2 Mini · 16:9 · 720p · 5s

Browse examples

AI video examples: press play, then try the prompt

Four short films show text-to-video and image-to-video workflows with Seedance 2 Mini and MiniMax H3. Videos load when you choose to play them; use the player controls to enable available audio.

Text to video

A product in golden light

A slow camera move reveals reflections in glass and warm light across stone.

Seedance 2 Mini16:9720p5sSilent

Image to video

Bring a travel photo to life

A gentle push-in, subtle hair movement and drifting clouds animate the original embrace.

Seedance 2 Mini16:9720p5sSilent

Text to video

A watercolor forest walk

An illustrated character walks through a sunlit forest with softly glowing spirits.

MiniMax H316:9768P5sAudio included

Text to video

A graphic product vignette

A stylized perfume film explores a different visual direction from the same product category.

MiniMax H316:9768P5sAudio included

Each example includes the prompt and settings used to create it. Use a prompt as a starting point; results vary. Reference-based examples need an uploaded image. Review the currently available model, settings and credits before generating.

Create videos from text or images

Describe a moving scene

Text to video starts with your description of a subject, action and camera movement. It is useful for a new product scene, an imagined environment or a short animation. Keep the scene focused enough for the selected duration.

Animate a photograph

Image to video starts from an uploaded image. Describe how the scene should move: a slight breeze, slow cloud movement or a restrained camera push-in. Say which parts should remain stable, especially faces, product shapes and the original composition.

The original couple sunset photograph, cropped to a 16:9 frame.
View the reference image

How to generate an AI video

  1. 1

    Choose a starting point

    Select Text to Video or Image to Video. For a photo animation, upload the first frame.

  2. 2

    Describe the motion

    Write the subject’s action and the camera movement. You can start with one of the example prompts.

  3. 3

    Set the clip options

    Choose a model, then open the settings panel for duration, ratio, resolution and any supported audio controls. Review CR.

  4. 4

    Generate and watch

    Sign in to start the task. Follow its status in your workspace, then play and download the finished video.

Match the video settings to your scene

Seedance 2 Mini is the default starting point. The model picker shows the current catalog, and each model determines which modes, inputs and settings are available.

Explore video settings
Duration and resolution
A five-second clip works well for one clear action or a gentle camera move. Choose from the duration and resolution options offered by the selected model; longer or larger outputs can change the credit amount.
Frame shape and source media
Use landscape for wide scenes and portrait for vertical layouts when the model supports them. Some modes inherit the source media’s ratio. Check the input requirements before uploading.
Audio and additional controls
Some models include audio, some offer an audio switch, and others produce silent video. Additional model-specific controls are under Advanced settings, alongside the available clip settings.

Give your video prompt a clear motion plan

A useful video prompt describes the subject, its action, the camera and what changes over time. The examples above include their full prompts so you can inspect the choices behind each clip.

One shot, one main action

For a short clip, choose one action that can read clearly. A slow turn, a light breeze or a gentle forward camera move is easier to follow than several scene changes.

Separate subject and camera motion

State whether the subject stays still while the camera moves, or whether the camera is fixed while the subject moves. Avoid contradictory instructions.

Keep references consistent

When animating an image, name the details that should stay stable. Use restrained motion for portraits and product shapes, then check the result before using it.

Review video credits for your selected settings

Video CR can change with the model, resolution, duration, generation mode and required source media. The server calculates the quote for your current configuration; changing a parameter updates the amount.

Usage-based tasks reserve the maximum shown in the quote, then return unused credits after final billing. Some modes need the source media before a quote can be calculated. Review the quote before generating and the billing status once the result is ready.

View credit plans

AI video generator FAQ

What is an AI video generator?

It creates a video from a text description or supplied media. This page supports text-to-video and image-to-video, with additional modes available according to the selected model.

Can I animate my own photo?

Yes. Select a model with Image to Video, upload a first-frame image and describe the motion you want. The sunset example shows a photo animated with a gentle camera move and small environmental changes.

Which model should I start with?

Seedance 2 Mini is the default. This page also includes MiniMax H3 examples so you can compare different visual results. Available modes and settings are shown in the generator for each model.

Can I generate a five-second video?

Yes, when five seconds is supported by the chosen model and mode. The examples here were requested at five seconds. Encoded files can include a small additional fraction of a second.

Can I create portrait and landscape videos?

Choose an available aspect ratio in the workbench. Supported ratios vary by model and mode; some workflows use the source media’s dimensions instead of a separate ratio setting.

Will the video include sound?

That depends on the model. The example labels show whether each clip contains audio. Examples start muted; use the player controls to listen. The generation settings show any available audio switch.

Does selecting an example start generation?

No. It fills the prompt and applies settings supported by the current catalog. For image-to-video you still need to upload a reference image. Review the settings and credits, then press Generate to submit.

Can I leave while the video is processing?

After submission, the task continues in the background. Return to My Works to check progress and open the result. If billing is still pending, the work’s status explains the current reservation or final charge.