archprompt.art

Example Prompts

Gallery of production-ready prompts.

History

Your recent generated prompts (saved locally in this browser).

Library (★ Favorites)

Your saved favorite prompts (local to this browser).

Import JSON

Tips & Guides

How to Think in V1.3

Start with Representation Type — decide what you want to produce (plan, section, diagram, detail, board) before thinking about style.
Use Representation Sub-Type to be specific (e.g. “Structural Node” vs “Technical Section”).
Let the tool guide you — prompts are automatically structured based on architectural intent.

Guidance & Preset Locking

You’ll see Guidance cards explaining what the selected representation is best used for.
If a field is locked, it’s intentional — changing it would usually reduce clarity for that representation.
Locked fields help beginners avoid common mistakes (for example: people in technical details).
Advanced users can still work fast — defaults only apply when a field is empty.

Basic vs Image-to-Image

Use Basic Prompt when creating a new image from scratch.
Use Image-to-Image when editing drawings, renders, plans, or sections you already have.
Enable Preserve Geometry to keep proportions, structure, and camera intact.

Final Tip

archprompt works best when you describe what the drawing is, not how the AI should think. Clear intent produces cleaner results.

About

Architecture-first prompting for consistent AI visuals.

archprompt.art is a prompt operating system for architecture.

It helps architects, designers, and students translate architectural intent into structured, repeatable AI prompts — without having to think like engineers or prompt specialists.

The goal is not to generate images for you. The goal is to help you communicate design thinking clearly to AI tools, so the results are consistent, intentional, and architecturally credible.

Why archprompt exists

AI image tools are powerful, but they often reward vague descriptions, trial-and-error prompting, and “magic words” with little design logic. This can create visuals that look impressive, but lack clarity, control, and repeatability.

archprompt.art was created to bridge that gap. It introduces architectural structure into prompting — organizing prompts around familiar concepts such as intent, style, materiality, geometry, lighting, and composition. The result is a workflow that feels closer to studio practice, not experimentation.

How it works

archprompt.art provides structured prompt builders for text-to-image generation and image-to-image restyling. Instead of writing prompts from scratch, you assemble them through clear architectural categories, producing prompts that are easier to reuse, remain consistent across a project, and work across different AI platforms.

Model-agnostic by design — built to work with Nano Banana, GPT Image, Stable Diffusion, Midjourney, and future tools.

What archprompt is (and is not)

archprompt.art is:

• Architecture-first
• Tool-agnostic
• Focused on repeatability and control
• Calm, minimal, and intentional

archprompt.art is not:

• An image generator
• A prompt marketplace
• A trend-driven AI playground
• A replacement for design thinking

AI is treated as a visual collaborator, not a shortcut.

Who it’s for

archprompt.art is designed for practicing architects and designers, students learning visual communication, and studios exploring AI as part of their workflow. Whether you’re producing concept visuals, atmospheric studies, diagrams, or marketing imagery, the focus remains the same:

clear intent → structured input → reliable output

Created by

Created by Milca Severo — architect, urban planner, and visual artist. Built from real studio work and a desire to bring architectural rigor into an emerging visual medium.

Follow on Instagram →

Version note

archprompt.art is evolving. Current versions focus on clarity, structure, and trust — future releases expand toward project-based workflows and deeper consistency controls.

Prompt Generator Txt-to-Image

Prompt Generator Image-to-Image