Documentation

Image

Text-to-image, reference images, batch count, and image tools.

Image generation creates PNG, JPG, or WebP files on your canvas from text prompts and optional reference images.

Modes in the generate bar

Mode labelWhen it applies
Text → ImagePrompt only, no upstream image
Ref Image → ImageOne or more upstream images connected by arrow

Bragi infers the mode from your upstream inputs.

Batch count

Set Count to x2, x3, or x4 to run parallel variants from the same prompt. Results stack vertically beside the source node.

Image models

Enable models in Settings → Bragi Canvas → Models → Image Models. Registered image models include:

  • Nano Banana Pro
  • Nano Banana 2
  • GPT Image 2
  • GPT Image 2.5
  • Grok Imagine
  • Midjourney v8
  • Midjourney niji 7
  • Luma Uni-1
  • Seedream 5.0 Pro
  • Seedream 5.0 Flash
  • Seedream 5.0
  • Seedream 4.5

Each model shows its own parameters in the generate bar, such as aspect ratio, size, quality, or stylize.

GPT Image 2.5

GPT Image 2.5 is available through APIMart and, from plugin 1.43.0, SVRouter. Enable the model and choose its provider in settings. Use your SVRouter gateway token for SVRouter; the gateway manages the upstream APIMart connection.

Choose Flare for faster generation or Sunburst for editing precision. Both use the same model entry, support 1K, 2K, and 4K output, and accept up to 16 connected reference images. The model also offers aspect ratio and quality controls.

Image file node tools

Select an image file node to see these toolbar actions:

ToolPurpose
Annotate imageDraw marks on the image in-canvas
Split gridSplit a grid or contact-sheet image into separate tile cards
360° viewerOpen a panorama viewer for PNG, JPEG, or WebP panoramas
DownloadSave the file outside the vault
MarkToggle purple highlight

Multi-select images

Select two or more image nodes only. The toolbar shows Collage (Create collage). This combines the selected images into one canvas layout.

Seedance Asset ID (advanced)

For Seedance video workflows with Volcengine, BytePlus, or TokenRouter face references:

  1. Right-click an image node.
  2. Choose Set Seedance asset ID.
  3. Pick the provider and paste the asset ID from your provider console.
  4. A green dot on the thumbnail indicates an ID is bound.

Asset IDs are provider-specific. The same image can have separate IDs for different providers.

On this page