Video Generation

Verdent can generate a short video from a text prompt or animate an image. Ask for the video in chat and Verdent runs the video tool, waits for the result, and returns a playable preview in the conversation.

Generate a video

Describe the scene

Include the subject, movement, camera direction, lighting, style, and intended use. For example: “Create a 6-second 16:9 product reveal with a slow camera orbit, soft studio lighting, and synchronized ambient audio.”

Add an image when needed

To create image-to-video content, provide a reachable image and explain whether it should act as the first frame, last frame, or a visual reference.

Wait for generation

Video generation usually takes several minutes. The response shows Generating video… until the final asset is ready.

Preview or download

The result plays as a muted loop in the message. Select the video to open a larger player with playback controls, or hover over it and select Download.

Available controls

You can ask Verdent to set:

duration from 4 to 15 seconds, or use the model default
480p, 720p, or 1080p resolution
16:9, 4:3, 1:1, 3:4, 9:16, 21:9, or adaptive aspect ratio
synchronized audio
watermark behavior
a repeatable seed
an optional final-frame image

The fast Seedance model does not support 1080p. Ask Verdent to use the standard model when you need 1080p output.

Output behavior

Verdent returns the final video URL and attempts to save an MP4 copy into the project. When requested, it also returns and saves the final frame as an image. If the local download fails but the hosted result is available, Verdent can still show the video from its URL.

Video previews work for local, Cloud, and SSH Remote projects through the appropriate project file or preview route.

Prompt tips

Describe motion over time, not only the opening frame.
Name the camera behavior: static, pan, orbit, tracking, handheld, or zoom.
Specify orientation for the destination, such as 16:9 for a presentation or 9:16 for a mobile feed.
State whether audio should be generated.
For image-to-video, explain what must remain visually consistent with the source image.