In the competitive world of architectural pitching, a breathtaking design concept is only half the battle — the other half is the presentation. How you sequence, frame, and narrate your visual ideas often determines whether you win or lose the bid.
For years, Adobe InDesign was the undisputed king of architectural presentations. But in the era of Artificial Intelligence, where concepts are generated and iterated upon at lightning speed, InDesign’s rigid, offline workflow has become a bottleneck.
Enter the modern, high-speed pitching stack: Generative AI (like Nuit) paired with Figma.
In this guide, we will explore why exporting your AI-generated architectural concepts into Figma is the ultimate workflow for creating winning, collaborative, and interactive client moodboards.
Why Figma for Architectural Presentations?
If you are a traditional architect, you might associate Figma purely with UI/UX designers building mobile apps. However, over the past three years, forward-thinking architecture studios have been migrating their presentation workflows to Figma en masse.
Why do architects use Figma? Architects use Figma because it is a cloud-native, infinitely collaborative vector design tool. Unlike InDesign, multiple team members can edit a presentation simultaneously. Its “Auto Layout” feature makes rearranging architectural moodboards and rendering grids completely frictionless, allowing for real-time presentation updates during client meetings.
When you combine Figma’s layout speed with the generation speed of an AI tool, you create a workflow that allows you to build a comprehensive, 20-page pitch deck in a single afternoon.
Nuit + Figma: The Perfect Synergy
A common question arises: If Nuit has an infinite canvas, why do I need Figma? Or conversely, If Figma has an infinite canvas, why do I need Nuit?
The answer lies in understanding the difference between the Workbench and the Gallery.
Nuit is the Workbench
The infinite canvas in an AI generator like Nuit is your messy, creative, exploratory workbench. This is where you write prompts, branch ideas, test a Brutalist concrete facade against a Mediterranean stucco facade, and generate 50 different variations. It is a visual record of your creative process. You do not want the client to see all 50 messy, experimental branches. This is the same branching design exploration technique that keeps your geometry stable while you test directions.
Figma is the Gallery
Figma is where you curate the chaos. It is the clean, structured, branded environment where you take the 4 best concepts from your Nuit workbench and arrange them into a compelling narrative. Figma is where you add your studio’s logo, precise typography, color hex codes, and project descriptions.
They do not compete; they synergize perfectly. Nuit generates the art; Figma builds the frame.
Step-by-Step: The AI to Figma Workflow
Let’s walk through the professional workflow for taking an AI concept and turning it into a final client presentation.
Step 1: Ideation and Curation (In Nuit)
Start on the Nuit canvas. Generate your massing studies, refine your materials using branching, and finalize your lighting. Once you have explored all your options, switch to “curation mode.” Identify the specific images that tell a cohesive story. You want:
- 1 “Hero” exterior shot.
- 2-3 Alternative exterior materials.
- 1 Masterplan/aerial view.
- 2 Interior mood shots. Export these specific high-resolution images to your local drive.
Step 2: Structuring the Moodboard (In Figma)
Open your studio’s Figma presentation template.
- The Grid: Drop your AI images into Figma. Use Figma’s
Auto Layoutfeature to arrange your images into a clean, masonry-style moodboard. Auto Layout ensures that if you decide to delete an image later, the rest of the board automatically snaps together perfectly without manual realignment. - Adding Reality: An AI image alone can feel untethered. Anchor it to reality by placing real-world material swatches (photos of actual physical concrete, timber, or fabric) next to the AI renders.
- Typography: Add elegant typography to label your concepts. (e.g., Option A: Raw Board-Formed Concrete | Option B: Fluted Travertine).
Avoid the “Image Dump.” The biggest mistake architects make with AI is generating 20 images and putting all 20 on a single slide. This overwhelms the client. Use Figma to ruthlessly curate your AI outputs. One massive, beautiful AI hero image surrounded by white space is infinitely more powerful than a grid of 12 mediocre iterations.
Step 3: Interactive Prototyping (The “Wow” Factor)
This is where Figma truly elevates the AI presentation. Because Figma is fundamentally a prototyping tool, you can make your architectural presentation interactive.
Imagine you used Nuit to generate an exterior building, and then used branching to generate two variations: a daytime shot and a nighttime shot.
In Figma, you can place these two images on top of each other and use the Smart Animate prototyping feature. During the client presentation, you can click a button, and the daytime AI render smoothly dissolves into the nighttime AI render.
This level of interactivity makes your AI concepts feel like a polished digital product, rather than just a static PDF. If you want to go further on presenting cleanly, see how to run AI client presentations without slop.
Collaborative Client Feedback
The final advantage of this workflow happens after the presentation.
Instead of exporting a massive 50MB PDF and emailing it to the client, you simply send them a view-only Figma link. The client can log in, pan across your beautiful presentation, and use Figma’s commenting tool to drop pins directly on the AI images. They might drop a pin on a window and comment: “Can we see this with darker framing?”
You take that feedback, jump back onto your Nuit workbench, branch the original image, apply the darker framing, export the new image, and drop it back into Figma. The client’s link automatically updates.
Conclusion: The Modern Pitch Stack
In 2026, the firms winning the best commissions are not just the ones with the best ideas; they are the ones who can communicate those ideas fastest and most beautifully.
By using generative AI to do the heavy lifting of spatial visualization, and Figma to do the heavy lifting of curation and layout, you eliminate the bottlenecks of traditional software. You move from a slow, manual rendering pipeline into a hyper-fast, highly collaborative design loop. The Nuit + Figma workflow isn’t just a time-saver; it is a competitive advantage. For structuring the boards themselves, see moodboards with sections for AI workflows.
Related reading
- AI Client Presentations Without Slop — Curate ruthlessly so your concept reads as intentional, not auto-generated…
- Moodboards With Sections for AI Workflows — Structure a moodboard so each section feeds the AI a clear material direction…
- Present Multiple Design Directions — Show the client a tree of options instead of one rigid building…
- Branching Design Exploration Technique — Test materials and lighting without losing your core geometry…
Frequently Asked Questions
Why use Figma for architectural moodboards instead of InDesign?
Figma is cloud-native and collaborative — multiple team members can edit the same presentation simultaneously, and clients can comment directly on the board. Its Auto Layout feature makes rearranging rendering grids frictionless, so you can update a moodboard live during a client meeting. InDesign’s offline, single-editor workflow can’t keep pace with AI-speed iteration.
If Nuit has an infinite canvas, why do I still need Figma?
Nuit’s canvas is your messy creative workbench — where you prompt, branch, and generate dozens of variations. Figma is the clean gallery where you curate the four best concepts into a branded narrative with typography, logos, and project descriptions. They don’t compete: Nuit generates the art, Figma builds the frame. You don’t want clients seeing all 50 experimental branches.
How do I move AI concepts from Nuit into Figma?
Curate the strongest images on your Nuit canvas — typically one hero exterior, a few material alternatives, a masterplan view, and a couple of interior shots. Export those at high resolution, then drop them into your Figma template and arrange them with Auto Layout. The whole transfer is a manual export-and-place, but it takes minutes.
Can I make an AI architectural presentation interactive in Figma?
Yes. Because Figma is a prototyping tool, you can layer two AI renders — say a daytime and nighttime version of the same building — and use Smart Animate so one dissolves into the other on click. This makes static concepts feel like a polished digital product rather than a flat PDF. It’s especially effective for showing lighting or material variations.
How do clients give feedback on a Figma moodboard?
Instead of emailing a heavy PDF, you send a view-only Figma link. Clients pan across the board and drop comment pins directly on the images — for example, pinning a window to ask for darker framing. You take that feedback back to your Nuit workbench, branch the image, regenerate, and drop the new version back into Figma, where the shared link updates automatically.
Does AI replace the architect’s design judgment in this workflow?
No. AI accelerates visualization and Figma accelerates layout, but curation, narrative, and material judgment stay human. The biggest mistake is dumping 20 generated images onto one slide — Figma’s value is helping you ruthlessly cut down to the few images that tell a cohesive story. AI output is concept-stage, not buildable documentation.
Try Nuit free — 100 credits, no card required. Generate and branch your architectural concepts in Nuit, then curate the best into a winning Figma moodboard. Start your project →