Model Selection

Model selection happens inside generative nodes. The selected model controls what the node can do and what settings appear.

Where model selection appears

Model selection is available on:

• Image Generation
• Video Generation
• Audio
• Music Generation

What changes when you switch a model

Changing models can affect:

• supported input types
• required source inputs
• available configuration fields
• cost per run
• estimated generation time

That means model choice is not just a quality preference. It can change how the workflow is wired.

How to choose a model

When picking a model, think about:

• media type you want
• whether you are starting from text or source media
• speed vs quality
• available credits
• provider-specific behavior

Model categories in practice

Image models

Best for:

• still image generation
• image editing
• upscaling
• character or style-driven visual creation

Video models

Best for:

• text-to-video
• image-to-video
• extension and upscale utilities
• advanced provider-specific workflows

Audio and music models

Best for:

• text to speech
• sound effects
• dialogue generation
• music generation
• audio utility tasks

Important rule

Always confirm the selected model matches the inputs already connected to the node.

Examples:

• a text-only model may ignore image inputs
• an image-to-video model may require an uploaded image
• a utility model may require upstream source metadata instead of a plain media file

Best practice

• Pick the model first when starting a new node.
• Then review the fields that appear.
• Then connect the inputs the model expects.

For pricing details, read Credits and Pricing.