alphaBell APPLIED MODEL LAB Browser only. Your draft stays here.

A SMALL TOOL FOR A BIG IMAGE MODEL

MOSS Layout Studio.

Put the words where you want them. Draw a composition, then export the exact layout prompt for MOSS-Image-Base.

01 / COMPOSE

A layout sketch, not a generated image. MOSS-Image-Base layout control is experimental.

Read the official format ↗Blind-check two generated images ↗
1024 × 1024 px
Text Object CharacterSchematic preview

You place boxes from the top left.
We convert to centered [y₁, x₁, y₂, x₂].

Ready to compose.

03

Take it to MOSS.

Save the editable draft, or copy the compiled JSON into MOSS-Image-Base’s prompt field. Set the same image dimensions in your generation tool.

Drafts stay in memory. Download yours before closing this page.

No accountNo inferenceNo uploads

Centered pixels · yxyx · image dimensions stay in your generation settings

Load the draft in MOSS Layout Prompt for ComfyUI at the file's width and height.

1

Compose in pixels

Drag and resize boxes. Write the exact visible words and describe each element’s appearance.

2

Copy the model prompt

The editor compiles absolute xyxy boxes into MOSS’s centered yxyx format. Dimensions are omitted from the prompt.

3

Generate, then compare

Use the same canvas dimensions in MOSS-Image. Blind-check two outputs against the target layout.