Back to Nodes
Kandinsky

Kandinsky

Official

Kandinsky text-to-video with resolution, length, and inference-step controls.

Nodespell AI
AI / Video / Open Source

Kandinsky text-to-video with resolution, length, and inference-step controls.

Kandinsky is an open text-to-video option with prompt, resolution, video length, and inference steps. It is useful when you want a straightforward controllable generation surface.

Use Kandinsky Distill when speed and a lighter setup matter more than exposing inference steps.

Inputs (1)

Prompt

String

The text prompt to guide video generation.

Multi InputMin: 0Max: 100
Parameters (4)

Prompt

String

The text prompt to guide video generation.

Default:

Resolution

String

Resolution of the generated video in W:H format. One of (768x512, 512x768, 512x512).

Default: 768x512

Num Inference Steps

Number

The number of inference steps.

Default: 30

video_length

String

Specifies the total length of the generated video in seconds.

Default: 5s
Outputs (1)

Output

Inferred

Output

Nodespell Team

Creator profile

Type

Node

Status

Official

Package

Nodespell AI

Category

AI / Video / Open Source

Input

Text

Output

Video

Keywords

Video GenerationAspect ControlResolution ControlLength Control
Use in Workflow