Answer

Short answer:

Rendair AI supports standard image formats including JPEG, PNG, WEBP, TIFF, HEIC, and AVIF, as well as single-page PDFs. For specific object insertion or modification workflows, the platform also accepts 3D file formats such as OBJ, FBX, STL, STEP, IGES, IFC, and 3DS.

Overview

Rendair AI is built to reduce friction in the design process by accepting the files architects and designers already use.

Rather than requiring heavy native project files (like .rvt or .pln) for every task, the platform primarily utilizes visual inputs, such as screenshots, exports, or sketches, to generate renderings. This approach allows for faster iteration and broader compatibility across different design software.

For most rendering tasks, a high-quality screenshot of your 3D viewport or a direct 2D export is the most efficient input. However, specific tools within the platform that handle object manipulation or insertion allow for the direct upload of 3D geometry.

Supported file categories

The platform distinguishes between visual inputs used for rendering and geometric inputs used for specific editing tasks.

Image and document formats

These formats are used for Image-to-Render, Sketch-to-Render, and general editing tools.

• Standard images: JPEG, JPG, PNG, WEBP.

• High-fidelity images: TIFF (useful for preserving detail before upscaling).

• Modern web/mobile formats: HEIC, AVIF.

• Documents: Single-page PDF.

3D model formats

These formats are supported specifically for tools that involve inserting or modifying 3D elements within a scene.

• Exchange formats: OBJ, FBX, STL, 3DS.

• CAD/BIM formats: STEP, IGES, IFC.

Inputs and outputs specifications

Understanding the technical constraints ensures predictable results when uploading files.

Input constraints

• Maximum file size: 16MB for image uploads.

• PDF handling: Must be single-page. If a PDF cannot be processed or edited, converting it to a high-quality JPG is recommended.

• 3D workflow: For full scene rendering, use a screenshot of the 3D model (textured or untextured) rather than the 3D file itself.

Output characteristics

• Aspect ratio: For text-to-render, supported ratios are 1:1, 2:3, 4:5, and 16:9. For all other tools, the output inherits the aspect ratio of the uploaded base image. For generations with base image, maximum aspect ratio size is between 1:2.5 and 2.5:1.

• Resolution: Default generations are approximately 1MP (or 2K for specific models like Nano Banana Pro).

• Upscaling: Outputs can be upscaled to 2K, 4K, 6K, or 8K depending on the specific tool and subscription plan.

When to use different formats

Choosing the right format depends on the stage of the design process.

• Screenshots (JPG/PNG): Best for "3D base to render" workflows. Capture the viewport from Revit, SketchUp, or Rhino to quickly generate realistic materials and lighting.

• Sketches (JPG/HEIC): Ideal for early concept validation. Photos of hand-drawn sketches or iPad exports work well here.

• Elevations (PDF/PNG): Use these for 2D-to-3D visualization to show clients how a flat drawing translates to a finished facade.

• 3D Files (OBJ/FBX): Use these only when the specific tool requires geometry for precise object placement or modification.

Limitations or notes

• File size management: If your image exceeds 16MB, compress it slightly or crop unnecessary borders before uploading.

• PDF editing: While PDFs are supported, they can sometimes behave unexpectedly depending on how they were exported. Rasterizing them to PNG often provides the most stable results for AI interpretation.

• 3D file usage: Uploading a 3D file (like an OBJ) does not render the entire scene automatically. The primary rendering engine is visual-based, meaning it interprets pixel data (screenshots) rather than geometry data. Maximum 3D file size is 300MB.