← Back to Documentation Home

Interior Design AI Models - Complete List

📊 Model Inventory

Text-to-Image Models (7 models)

Used when user provides only a text prompt (no reference image)

Model ID Name Provider Status
flux-dev FLUX.1-dev Replicate ✅ Working
flux-schnell FLUX.1-schnell Replicate ✅ Working
sdxl SDXL Replicate ✅ Working
playground-v2.5 Playground v2.5 Replicate ✅ Working
stable-diffusion-3 Stable Diffusion 3 Replicate ✅ Working
kandinsky-2.2 Kandinsky 2.2 Replicate ✅ Working
proteus-v0.2 Proteus v0.2 Replicate ✅ Working

Image-to-Image Models (7 models)

Used when user provides reference image + prompt for interior transformation

Model ID Name Provider Status
jschoormans/comfyui-interior-remodel ComfyUI Interior Remodel Replicate ✅ Working
julian-at/interiorly-gen1-dev Interiorly Gen1 Dev Replicate ✅ Working
davisbrown/designer-architecture Designer Architecture Replicate ✅ Working
erayyavuz/interior-ai Interior AI Replicate ❌ Failing
jschoormans/interior-v2 Interior V2 Replicate ❌ Failing
adirik/interior-design Adirik Interior Design Replicate ❌ Failing
rocketdigitalai/interior-design-sdxl Interior Design SDXL Replicate ❌ Failing

🎯 Smart Model Selection Logic

Scenario 1: Text-to-Image Request

When the user provides only a text prompt (e.g., "Generate a modern minimalist bedroom with oak flooring"), all 7 text-to-image models are used: FLUX.1-dev, FLUX.1-schnell, SDXL, Playground v2.5, Stable Diffusion 3, Kandinsky 2.2, and Proteus v0.2. This produces 7 different variations.

Scenario 2: Image-to-Image Request

When the user provides a reference image along with a prompt (e.g., "Transform this into a modern minimalist style"), only the 3 working image-to-image models are used: ComfyUI Interior Remodel, Interiorly Gen1 Dev, and Designer Architecture. This produces 3 transformed variations.

Scenario 3: Custom Model Selection

When the user specifies particular models in the request (e.g., "models": ["flux-dev", "sdxl"]), only those specified models are used, producing the corresponding number of variations.

🔧 API Behavior

Default Behavior (No Models Specified)

When no models are specified in the request:

Custom Model Selection

When a models array is provided in the request, the API filters the full model list to include only those whose IDs appear in the request array.

📈 Expected Results

Request Type Models Used Expected Images
Text-to-image (default) 7 7 variations
Image-to-image (default) 3 3 transformations
Custom selection Variable As specified

✅ Complete Model Count: 14 Total