Davis Unveils Its Gaudi-1 Model for Automated Design
Edited by Adam Harrie — May 11, 2026 — Tech
This article was written with the assistance of AI.
References: meetdavis & eu-startups
Davis, a Paris-based AI-native real estate startup, launched Gaudi-1, a proprietary model for automated architectural generation designed to produce architect-grade designs under regulatory constraints. The company paired the debut with a €4.6 million pre-seed round led by Heartcore Capital and Balderton Capital, featuring an approach that generates buildings as structured compositions of elements such as rooms, walls and layouts.
Gaudi-1 uses a discrete generative method rather than pixel-based diffusion, enabling outputs that encode volumetrics, floor plans and space planning while incorporating regulatory, technical and market constraints. Davis said human architects review each output, and the model achieved top scores on floor-plan benchmarks including RPLAN and MSD across IoU, FID and KID metrics.
For developers and investors, Gaudi-1 promises faster feasibility and concept stages by compressing timelines from months to days while improving iteration speed and regulatory compliance. Delivered as a service rather than standalone software, the system targets multiple asset classes and geographies while keeping architects in the loop.
Image Credit: Davis
Gaudi-1 uses a discrete generative method rather than pixel-based diffusion, enabling outputs that encode volumetrics, floor plans and space planning while incorporating regulatory, technical and market constraints. Davis said human architects review each output, and the model achieved top scores on floor-plan benchmarks including RPLAN and MSD across IoU, FID and KID metrics.
For developers and investors, Gaudi-1 promises faster feasibility and concept stages by compressing timelines from months to days while improving iteration speed and regulatory compliance. Delivered as a service rather than standalone software, the system targets multiple asset classes and geographies while keeping architects in the loop.
Image Credit: Davis
Would you use an AI design canvas for floor plans?
Helps decide what AI-for-architecture content to prioritize and which product features and adoption barriers matter most to readers considering new design tools.
1 / 3
When was the last time you tried a new design tool for work?
2 / 3
If you were choosing tools, would you try AI that drafts floor plans from text?
3 / 3
Which AI feature would make you most likely to switch design tools?
Trend Themes
-
Discrete-model Architectural Generation — Potential to produce architect-grade volumetric and floor-plan outputs that map directly to building elements and technical constraints.
-
Regulatory-constrained Generative Models — Models that encode zoning, code and market constraints directly into design outputs, yielding higher first-pass compliance rates.
-
Design-as-a-service Platforms — Delivery as a service centralizes rule-set updates and multi-geography capabilities, shifting value toward scalable design pipelines rather than standalone software.
Industry Implications
-
Real Estate Development — Faster feasibility and concept iterations that can shorten holding periods and accelerate deal velocity for mixed-use and residential portfolios.
-
Architecture Firms — Human-AI collaboration approaches that preserve architect oversight while enabling higher-volume concept generation and more strategic client advisory work.
-
Construction and Modular Builders — Structurally explicit, element-based outputs that can be translated into panelized or modular components to streamline prefabrication and supply-chain coordination.
6.3
Score
Popularity
Activity
Freshness
