Specialized Visualization for AEC

Fenestra is a web-based rendering platform designed to bridge the gap between architectural conceptualization and final presentation. Founded in 2024, the platform addresses a specific bottleneck in the architecture, engineering, and construction (AEC) industry: the time and technical expertise required to produce high-fidelity visualizations. Traditional workflows involve complex lighting setups and long rendering times in software like V-Ray or Enscape. Fenestra replaces this with an AI-driven pipeline that accepts sketches, elevations, and 3D files to generate imagery in seconds.

The platform is built on a collection of generative vision models that the company describes as "hand selected by architects for architects." By focusing exclusively on the Archviz domain, Fenestra avoids the generic aesthetic often found in general-purpose image generators. This specialization is reflected in their user base, which includes designers from notable firms and institutions like Mikhail Riches, Shop Architects, and Cornell University.

The Multi-Model Pipeline

Rather than relying on a single foundation model, Fenestra utilizes a modular stack of models for different stages of the design process. The platform features several distinct modes:

• Create Mode: Transforms raw geometry or 2D CAD drawings into photorealistic renderings.
• Edit Mode: Uses natural language to modify existing designs, such as adding people, removing furniture, or changing facade materials.
• Enhance Mode: Sharpening low-resolution images into 4K visualizations.
• Animate Mode: Converting static images into cinematic flythroughs.

The underlying model selection includes several versions of Flux (Kontext Max, Pro, and Dev), Stable Diffusion XL, and specialized video models like Runway and Wan 2.2. A notable component is "Nano Banana," a proprietary-tuned model used for architectural diagrams and precise image editing. This variety allows users to choose between rapid iteration for concepts and high-fidelity output for client presentations.

Market Position and User Experience

Fenestra competes in a rapidly growing category of AI design tools. Unlike generic competitors like Midjourney or DALL-E, Fenestra supports spatial data formats essential to architectural workflows, including FBX, GLB, and Rhino (3DM) files. This allows architects to maintain the structural integrity of their designs while experimenting with materiality and lighting.

The platform operates on a credit-based subscription model. Tiers range from an Individual plan designed for students and solo designers to a Studio plan for high-volume firms. Enterprise options add team management features and shared credit pools. Data privacy is a key focus for their professional users; Fenestra states that custom-trained models and uploaded project data are stored securely and not shared with third parties or used for general model training.

While the company is relatively new, it has already recorded over 100,000 generations across 11,000 unique projects. The leadership, spearheaded by Shaun McCallum, emphasizes that the tool is intended to enhance rather than replace the architect, focusing on the "advantage" gained by designers who can iterate faster than traditional peers. The platform's ability to handle 2D architectural elevations and 3D interactive flythroughs positions it as a specialized companion to existing CAD software.