Back
AI Machine Learning Deep Learning Computer Vision NLP Linguistic Processing Image Processing
01234567/7

AI: Broad field of intelligent systems.

Machine Learning: Algorithms that learn patterns from data.

Deep Learning: Multi-layer neural networks.

Computer Vision: Understanding images & video.

NLP: Understanding & generating text.

Linguistic Processing: Rules & classical analysis.

Image Processing: Filtering & transforms.

Generative AI Tools Cheatsheet 2025 — Compact

📝 Text Generation & Language Models

ChatGPT

OpenAI

Claude

Anthropic

Gemini

Google

Copilot

Microsoft

Perplexity

Perplexity AI

Jasper

Jasper AI

Deepseek

Deepseek

🎨 Image Generation & Visual AI

Midjourney

Midjourney Inc

Stable Diffusion

Stability AI

ComfyUI

Community

DALL·E 3

OpenAI

Imagen

Google

Adobe Firefly

Adobe

Leonardo AI

Leonardo.ai

Runway ML

Runway

🎬 Video Generation & Animation

Sora

OpenAI

Runway Gen-2

Runway

Pika

Pika Labs

Stable Video

Stability AI
NEW

Banana

Google

🎵 Audio & Music Generation

Suno

Suno

Udio

Udio

ElevenLabs

ElevenLabs

Mubert

Mubert

Boomy

Boomy

💻 Code Generation & Development

GitHub Copilot

GitHub / OpenAI

Cursor

Cursor

Huggingface

Huggingface

Codeium

Codeium

Amazon CodeWhisperer

Amazon

🛠️ Specialized & Workflow Tools

Zapier AI

Zapier

Notion AI

Notion

Figma AI

Figma

Canva AI

Canva

Otter.ai

Otter.ai

Cheatsheet 2025 — compact view with live icons

AI as the Middleware of Design Workflows

Coding flows into AI → AI orchestrates outputs to Fabrication, Visualisation, and BIM.

Coding
AI Middleware (Gen AI Orchestrator)
Fabrication
Visualisation
BIM
Generative AI for Architecture — Cheatsheet 2025

✍️ Text → Text (Concepts, Notes, Coding)

ChatGPT

OpenAI

Claude

Anthropic

Gemini

Google
RESEARCH

Perplexity Pro

Search + Codes

🖼️ Text → Image (Plans, Renders, Concepts)

Stable Diffusion XL

Stability AI

ComfyUI Workflows

Community
ARCHVIZ

Midjourney

Midjourney

DALL·E 3

OpenAI

Leonardo AI

Leonardo.ai

🎨 Image → Image (Sketch → Render, Plan → 3D)

PLAN → 3D

ComfyUI + ControlNet

Community
REFINE

Runway Img2Img

Runway

Stable Diffusion Img2Img

Stability AI
MESH

Luma / NeRF

Luma AI

🧱 Text → 3D / CAD / BIM (Massing, Sites, Feasibility)

URBAN

Autodesk Forma

Autodesk
SCHEMES

Hypar

Hypar
FEASIBILITY

TestFit

TestFit Inc.
PARAM

Finch

Finch3D
API

KittyCAD

CAD via API
XR

Arkio

Collaborative Massing
GEN PLAN

Hypar Rooms

Program → Layout
Rhino.Inside

Rhino / Grasshopper

Procedural BIM

Generative AI for Architecture — 2025 Cheatsheet (Text→Text, Text→Image, Image→Image, Text→3D/CAD/BIM)

🌀 AI Diffusion Models — Overview

Art Generation Image Generation + Editing Fine-Tuning Models Controlled Generation Realtime Editing Midjourney DALL·E LoRA ControlNet KREA Diffusion Models look .ai mmm.ai

Automation

Predefined, rule-based tasks executed automatically

  • CoreBoolean logic
  • TasksDeterministic, predefined
  • StrengthsReliable • Fast
  • WeaknessesCan’t adapt • Limited scope
  • ExampleNotify Slack when a lead signs up

AI workflow

Pipeline that calls an LLM/API for one or more steps

  • CoreBoolean + fuzzy logic
  • TasksDeterministic with flexibility
  • StrengthsHandles complexity • Pattern recognition
  • WeaknessesNeeds data • Harder to debug
  • ExampleScore & route inbound leads with ChatGPT

AI agent

Autonomous program for non-deterministic tasks

  • CoreFuzzy logic + autonomy
  • TasksAdaptive, non-deterministic
  • StrengthsHighly adaptive • Human-like behavior
  • WeaknessesLess predictable • Slower
  • ExampleResearch the web & update CRM automatically
Concept Development

Key Actions

  • Moodboards & concept art
  • Visual exploration & style transfer

AI Techniques

  • Diffusion models
  • Prompt→Image
  • ControlNet guidance

Tools (ex.)

  • Midjourney
  • DALL·E
  • Krea
  • Stable Diffusion

Output

Concept visuals & moodboards

Schematic Design

Key Actions

  • Layout options & massing
  • Adjacency & circulation
  • Parametric exploration

AI Techniques

  • Generative design
  • Optimization algorithms

Tools (ex.)

  • Autodesk Generative Design
  • Finch
  • PlanFinder

Output

Schematic layouts & zoning schemes

Design Development

Key Actions

  • Material & structural analysis
  • Render & animation

AI Techniques

  • CNNs for detection
  • Generative rendering

Tools (ex.)

  • Lumion (AI)
  • NVIDIA Canvas
  • Runway · Pika · Sora

Output

Detailed visuals & render videos

Simulation

Key Actions

  • Energy modeling
  • Daylight & climate analysis
  • Structural testing

AI Techniques

  • ML-assisted simulation
  • Predictive analytics

Tools (ex.)

  • Ladybug / Honeybee
  • Karamba
  • TopologicPy

Output

Performance reports, validated design

Construction

Key Actions

  • Digital twin integration
  • Robotics & prefabrication
  • Site monitoring

AI Techniques

  • Computer vision on site
  • Robotics planning

Tools (ex.)

  • Bentley Synchro
  • Boston Dynamics Spot
  • Smart Construction AI

Output

Efficient prefabrication, tracked progress

Building Management

Key Actions

  • Energy optimization
  • Predictive maintenance
  • Occupant comfort

AI Techniques

  • IoT + AI monitoring
  • Forecasting & anomaly detection

Tools (ex.)

  • Digital Twin platforms
  • BIM 360 AI
  • Honeywell Forge

Output

Smart, sustainable operations

Tool Category Link
Imerso
As-built / Site monitoring
Construction site monitoring Visit Imerso
Aurivus
Scan-to-BIM AI
Scan → BIM Visit Aurivus
Ark Design (ARK)
Schematic design / feasibility
Schematic AI / Feasibility Visit ARK
Skema
Schematic → BIM automation
Concept design → BIM Visit Skema
Add your tool
Category
Category Link

Source: AEC AI Hub (database) + vendor pages. See notes below for sources and how to replace logos.

ChatGPT for Architects

Smart help across design, BIM, visualization, scripting, and communication.

Design

Concept & Precedent Ideation

Generate concept directions, precedent lists, and option trees aligned with site, climate, and program.

Draft 5 concept narratives for a coastal cultural center in Doha; list key precedents and massing moves per option.
Grasshopper

Parametric Script Drafting

Turn natural language into GhPython/C# snippets with inputs/outputs, constraints, and notes.

Write GhPython to generate a seating bowl: inputs (radius, tiers, step rise), outputs (curves, seat points).
Revit/Dynamo

BIM Automation

Create Dynamo Python to place rooms, tag them, and flag doors < 80 cm with a warning marker.

Generate a Dynamo Python script to create rooms from polyloops, place tags, and mark doors <800 mm.
BIM QA

Model Checks & Reports

Ask for rule-based checks: naming, levels, view templates, fire egress, travel distances, etc.

Create a BIM QA checklist for a mixed-use tower: naming rules, views, sheets, egress, and export as table.
Viz

Image & Animation Prompts

Craft Midjourney/ComfyUI prompts, camera notes, and storyboard beats for short reels.

Write a Midjourney prompt for a winter plaza in Islamic Cairo; add camera & mood board notes.
Analysis

Site & Climate Briefs

Summarize climate data, sun paths, winds, & regulations; propose passive strategies.

Summarize site constraints for Lusail plot: setbacks, heights, wind/solar, and passive cooling strategies.
Code

Debug & Refactor

Paste errors; get minimal, clean fixes for Python (GH/Dynamo), JS widgets, or API calls.

Here’s my GhPython error trace; suggest a minimal fix and a cleaner function design.
Docs

Competition & Client Packs

Turn notes into polished briefs, executive summaries, and slide outlines (EN/AR).

Turn these notes into a 6-slide client deck with key visuals and talking points (Arabic + English).
Scheduling

Roadmaps & Checklists

Create phased Gantt-style task lists with dependencies for studio deliverables.

Plan a 6-week concept→DD roadmap with tasks, owners, and review gates.
Fabrication

G-code & Toolpaths (Intro)

Outline non-planar 3D-printing strategies; generate starter G-code patterns.

Explain non-planar G-code for a ribbed shell; generate a safe starter pattern with comments.
Research

Lit Reviews & Summaries

Get crisp summaries of papers, compare methods, and extract architecture takeaways.

Summarize this paper for architects: key findings, design implications, and limits.
QA

Code → Model Consistency

Cross-check design rules vs. model outputs; report violations in a table.

Compare these design rules to the exported model data; list violations + suggestions.
Prompt copied ✓
Images

Text → Image

High-fidelity architectural exteriors with materials, lighting, climate context, and composition notes.

Create a dusk exterior of a contemporary mosque in Doha, basalt + weathered steel, soft rain, wet ground reflections, 35mm lens, f/4, rule-of-thirds, hero shot.
Images

Interior Spaces

Render residential or commercial interiors with furniture, finishes, lighting scenarios, and atmosphere.

Luxury hotel lobby in Doha: travertine floors, brass chandeliers, deep green velvet seating, date palm in planter, morning light through mashrabiya screens, 24mm, eye-level, warm elegant mood.
Images

Material Close-Ups

Detail shots showing texture, grain, joinery, and material quality for presentations.

Macro shot of walnut wood panel junction with brass inlay detail, raking side light, shallow depth of field, 100mm lens, architectural detail photography style, warm tones.
Images

Landscape & Courtyards

Outdoor spaces with native planting, hardscape, water features, and regional climate considerations.

Desert courtyard garden in Qatar: ghaf trees, gravel paths, limestone walls, water feature with mist, dappled shade, golden hour, people walking, 35mm, Peter Walker landscape style.
Images

Lighting Studies

Compare dawn, midday, dusk scenarios to show daylighting performance and artificial lighting design.

Museum gallery at three times: dawn (cool blue wash), noon (bright skylight), dusk (warm accent spots). White walls, timber floor, sculptures, same camera angle, 28mm wide.
Images

Aerial & Context Shots

Bird's-eye views showing site integration, urban fabric, landscape, and neighborhood context.

Aerial view of mixed-use development in Lusail: curved towers, palm-lined boulevards, marina edge, late afternoon, drone perspective 45° angle, showing skyline context, contemporary Gulf urbanism.
Images

Vignette & Lifestyle

Editorial-style detail shots with styling, human activity, and intimate spatial moments.

Cozy reading nook: oak window bench with cushions, morning light through sheer curtain, ceramic mug on sill, book open, soft focus background, 50mm f/1.8, Kinfolk magazine aesthetic.
Images

Conceptual Mood Boards

Abstract explorations of light, shadow, texture, and spatial atmosphere for early design phases.

Abstract concept: filtered light through geometric screens, shadow patterns on textured plaster, warm amber to deep sepia tones, James Turrell meets Islamic geometry, ethereal contemplative mood.
Sora

Text → Video

Create cinematic walkthroughs with camera moves, environmental changes, and atmospheric storytelling.

Slow dolly push through a spa corridor: sandstone walls, indirect cove lighting, reflection pool, steam rising, palm shadows moving on wall, 24fps, cinematic 2.39:1, serene luxury mood.
Sora

Image → Video

Animate a still render with camera movement, people activity, or environmental effects.

Animate this plaza exterior: add gentle breeze moving palm fronds, people walking across frame, golden hour light slowly shifting, birds flying overhead, keep architecture static, 10-second loop.
Sora

Time-Lapse Sequences

Show spaces transforming through day-night cycles, seasons, or construction phases.

Time-lapse of courtyard from dawn to dusk: shadows rotating across limestone walls, people coming and going, fountain reflecting changing sky, 30 seconds compressed, fixed wide camera.
Sora

Spatial Experience Films

First-person or tracking shots showing circulation, scale, and experiential qualities of spaces.

First-person walk through museum: start in compressed dark entry, emerge into tall daylit atrium, camera tilts up revealing skylight, continue toward gallery, footsteps echoing, 15 seconds, cinematic pacing.

Visual Superpowers: Images & Video

Text→Image, Image→Image, Outpainting, Variations — and Sora video generation.

Images

Text → Image

High-fidelity concept images with clear materials, lighting, and composition notes.

Create a dusk exterior of a parametric museum in Doha, basalt + weathered steel, soft rain, wet ground reflections, 35mm lens, f/4, rule-of-thirds, hero shot.
Images

Image → Image (Edits)

Upload a base image + mask; replace sky, furniture, materials, or signage in place.

Edit the masked facade panels to perforated aluminum with Islamic geometric pattern, keep scale + perspective, match existing lighting.
Images

Outpainting (Extend)

Expand canvas to widen shots, add context, and keep style continuity.

Extend canvas 200% left/right; continue plaza paving, add palm rows, subtle crowd, maintain same lighting + camera height.
Images

Variations & Style Reference

Generate alternatives or lock a visual language by referencing a style image.

Generate 4 variations preserving massing + camera; shift to warm sandstone, bronze trims, evening sky; follow the attached style image for color grade.
Images

Text on Image (Plans/Boards)

Ask for crisp text labels, titles, or legend overlays directly in the render.

Render concept board with title “Coastal Cultural Center — Doha”, add 4 captions under images, font clean sans (board size 1920×1080).
Sora

Text → Video

Create cinematic camera moves, environment, and timing for short clips.

A slow dolly-in toward a desert pavilion at golden hour; wind moving tall grasses; 24fps; gentle handheld micro-jitter; 6–10s clip.
Sora

Image → Video

Animate a still render (clouds, foliage, water) while preserving composition.

Animate this plaza render: add light rain streaks, wet ground ripples, subtle crowd motion; keep lens + framing unchanged; 9s.
Sora

Remix / Extend

Blend two clips, extend a shot, or edit with a mask to change only parts.

Extend this corridor tracking shot by ~5s; keep same speed + exposure; replace ceiling lights via masked edit with linear skylights.
Sora

Shot Direction

Specify lens, move, framing, and pace for tighter control.

Architectural fly-through: 18mm ultra-wide, gimbal-smooth orbit 90° around pavilion, parallax foreground plants, 8s, natural audio muted.

Note: Availability, max duration, and resolution depend on your plan & current access.

Copied ✓

Examples from OpenAI Documentation

Real images shown in OpenAI’s articles — outpainting, styles, and compositing.

OpenAI: DALL·E Outpainting example (Vermeer expansion)
Outpainting Extend beyond borders while keeping lighting, texture & perspective consistent. Source ↗
OpenAI: DALL·E Outpainting collage example 1
Style Continuity Large canvas expansions with consistent palette & composition. Source ↗
OpenAI: DALL·E Outpainting collage example 2
Variations Multiple compositions derived from one core idea. Source ↗
OpenAI Sora — see video examples on the page
Sora See text→video & image→video clips on the Sora page. Source ↗

All images link to official OpenAI pages for full context and credits.

DALL·E Examples

Core functions: text→image, image→image (edits), outpainting, variations.

DALL·E text-to-image example
Text → Image Generate original renders directly from descriptive prompts. Docs ↗
DALL·E edit (masked area replaced)
Image → Image (Edits) Replace masked areas with new content while keeping rest intact. Docs ↗
DALL·E outpainting (Girl with a Pearl Earring expansion)
Outpainting Extend canvases beyond borders, keeping lighting & style consistent. Blog ↗
DALL·E variations (multiple outputs from one input)
Variations Generate multiple stylistic takes from one base image or prompt. Docs ↗

All images are official DALL·E examples from OpenAI’s documentation.

Custom GPT Categories

Architect-focused GPTs you can build: specialized assistants for design, BIM, visualization, and more.

Design

Concept Design GPT

Helps with massing studies, precedent analysis, and quick ideation for competitions or studios.

BIM

Revit & QA GPT

Checks model standards, flags issues (door width, room tags), and drafts Dynamo/Revit API scripts.

Visualization

Rendering & Prompt GPT

Crafts Midjourney/ComfyUI prompts, camera settings, or storyboard notes for architectural films.

Scripting

Parametric Code GPT

Generates clean GhPython, C#, or Dynamo scripts with inputs/outputs tailored to architectural workflows.

Fabrication

Maker Lab GPT

Produces CNC/3D printing G-code samples, panel unfolding strategies, or robotic toolpath notes.

Research

Academic GPT

Summarizes papers, compares algorithms, and translates technical content into architectural insights.

Docs

Client Pack GPT

Turns raw notes into polished briefs, bilingual reports, or pitch decks with clear visuals.

PM

Roadmap GPT

Creates task lists, phased Gantt outlines, and delivery checklists for teams and studios.

Workflow: Customize Your GPT

From idea → building → testing → sharing.

1. Define Goal 2. Add Knowledge 3. Configure Style 4. Test & Share
Viz

Architectural Exterior Renders

Generate detailed Midjourney prompts for building exteriors with climate, materials, lighting, and cultural context.

Write a Midjourney prompt for a contemporary mosque courtyard at golden hour in Doha; include palm shadows, mashrabiya patterns, limestone materiality, camera notes (low angle, 24mm), atmospheric haze, and mood references (Hassan Fathy meets Tadao Ando).
Viz

Interior Design & Visualization

Craft Midjourney prompts for interior spaces with furniture, finishes, lighting scenarios, and spatial atmosphere.

Create a Midjourney prompt for a luxury majlis interior with modern Gulf heritage fusion: brass mashrabiya screens, terrazzo flooring, jewel-tone velvet seating, warm indirect lighting. Specify camera (eye-level, 35mm), late afternoon sun, material details (polished marble, woven textiles), and mood (contemporary Qatari elegance meets boutique intimacy).
Viz

Material & Texture Studies

Generate close-up renders focusing on material details, textures, and surface quality for presentations.

Write a Midjourney prompt for a material detail study: travertine wall with raking light, showing natural veining, matte finish, macro photography (100mm), shallow depth of field, styled like architectural digest detail shots.
Viz

Lighting Scenario Renders

Illustrate different lighting conditions and their impact on spaces for daylighting and artificial lighting studies.

Generate a Midjourney prompt for a hotel lobby in three lighting scenarios: dawn (cool blue), midday (bright neutral), dusk (warm amber). Include pendant fixtures, natural light from skylights, camera (wide 18mm), architectural photography style.
Viz

Custom Furniture & Fixtures

Visualize bespoke furniture designs with joinery details, materials, and styling in context.

Create a Midjourney prompt for a custom walnut credenza with brass inlay details: show grain patterns, brushed metal hardware, styled with ceramic objects, studio lighting (softbox), product photography aesthetic, white seamless background.
Viz

Color Palette & Schemes

Develop mood boards showing color relationships, material combinations, and finish options for client presentations.

Write a Midjourney prompt for an interior color palette board: sage green walls, terracotta accents, natural oak, aged brass, cream textiles. Show material swatches arranged flat-lay style, even lighting, styled like Norm Architects mood boards.
Viz

Spatial Flow & Layouts

Visualize spatial sequences, circulation paths, and experiential journeys through buildings.

Generate a Midjourney prompt for a museum gallery sequence: entry transition from compressed to expansive, indirect skylights, polished concrete floors, white walls, human scale figures, cinematic wide shots (21mm), Ando-inspired spatial drama.
Viz

Detail Shots & Vignettes

Create intimate, editorial-style vignettes highlighting design details and material intersections.

Create a Midjourney prompt for an architectural detail vignette: corner window with oak frame, soft curtain filtering light, ceramic vase on sill, morning glow, shallow focus (50mm f/1.8), editorial lifestyle photography like Kinfolk magazine.
Viz

Landscape & Outdoor Spaces

Render plazas, gardens, courtyards with vegetation, hardscape, and climate-appropriate detailing.

Write a Midjourney prompt for a desert courtyard garden: native plantings (ghaf trees, succulents), decomposed granite paths, water feature with misting, dappled shade, late afternoon light, human activity, eye-level view (35mm), Peter Walker landscape aesthetic.
Viz

Renovation Before/After

Show transformation narratives for adaptive reuse and renovation projects with comparative views.

Generate a Midjourney prompt pair for warehouse-to-loft conversion: BEFORE (industrial shell, exposed structure, raw concrete), AFTER (polished floors, mezzanine, glazing, warm lighting), same camera angle (wide 24mm), dramatic transformation narrative.
Viz

Cultural & Regional Context

Embed designs in specific cultural and geographical settings with authentic contextual elements.

Create a Midjourney prompt for a Qatari family villa: contemporary Islamic architecture, wind tower reinterpreted, white limestone, lush courtyard with date palms, carved mashrabiya, golden hour, aerial view showing Doha skyline, regional authenticity.
Viz

Conceptual Mood Boards

Develop abstract concept images exploring form, light, materiality, and spatial atmosphere for early design phases.

Write a Midjourney prompt for a conceptual mood board exploring "filtered light + porosity": abstract mashrabiya shadows on textured surfaces, light and shadow play, tonal range (warm whites to deep amber), James Turrell meets Islamic geometry, ethereal and contemplative.