Send a photo and what to change. Get back a photorealistic render of that exact change, in that exact photo, in seconds. Everything you see moving here is unretouched output from the live API.
POST /public/v1/renovate
X-API-KEY: $RENOVATE_API_KEY
{
"image_url": "https://your-site.com/customer-room.jpg",
"guidance": "Repaint the kitchen cabinets bright white with matte black hardware. Keep the rest of the kitchen unchanged.",
"renovation_spectrum": "tweak"
}Model Context Protocol
Plug the API straight into your AI tools — Claude, ChatGPT, Codex, Cursor, or any MCP client.
claude mcp add --transport http renovateai https://mcp.renovateai.app/mcpRuns in your terminal with Claude Code.
The integration
Map your catalog once. Every render carries your exact SKU into the customer’s own room — the picture a shopper approves is the product that ships.
Model Context Protocol
Authorize the Renovate AI server once from Claude, ChatGPT, Codex, Cursor, or any MCP client, and call every endpoint straight from chat — renders, staging, upscales — each one drawing from your API credits.
Grab the install command for your client at the top of the page. One line — or one connector URL — registers Renovate AI.
Your client opens an OAuth consent — authorize once and your key and credits link automatically, no config files.
Ask it to renovate, stage, or upscale a photo — every call runs on the live API and draws from your credits.
ConnectedCatalog-exact
The materials array carries your SKU, texture image, and installation
location. The engine applies your product to the customer’s photo — one
schema across every surface.
Flooring, cabinetry, countertops, backsplash, paint, siding, and ceilings — map your catalog once, render everywhere.
{
"image_url": "https://your-site.com/customer-room.jpg",
"guidance": "Install the following flooring: Natural engineered hardwood. Keep the rest of the room unchanged.",
"renovation_spectrum": "tweak",
"materials": [{
"material_id": "fl-hwd-001",
"material_name": "Oak Hardwood",
"color_id": "fl-hwd-001-nat",
"color_name": "Natural",
"image_url": "https://your-cdn.com/materials/fl-hwd-001-nat.png",
"installation_location": "FLOORS",
"metadata": { "width": "7in", "finish": "Wire Brushed", "species": "White Oak" }
}]
}curl -X POST https://api.tech.renovateai.app/public/v1/renovate \
-H "Content-Type: application/json" \
-H "X-API-KEY: $RENOVATE_API_KEY" \
-d '{
"image_url": "https://your-site.com/customer-room.jpg",
"guidance": "Repaint the kitchen cabinets bright white with matte black hardware. Keep the rest of the kitchen unchanged.",
"renovation_spectrum": "tweak"
}'The full surface
/renovatePhotorealistic render of specific materials or changes applied to a customer photo. Drive it with guidance text, or attach a materials array — SKU, texture image, installation location — for catalog-exact results.
/virtual-stagingStage empty listings by room type and style. Project-based: create a project per listing, generate styled variations per room.
/kitchen_cabinetsPurpose-built cabinet refacing and repainting — the highest-volume surface in remodeling, tuned as its own endpoint.
/style_transferRe-render a space in a named design style while preserving its geometry.
/creative_upscalePrint- and marketing-grade enlargement of generated or original photos.
/analyzeStructured understanding of room photos — single and batch.
Complete request and response schemas live in the API reference.
In production
Let shoppers see your exact SKU — flooring, cabinets, countertops, siding — installed in their own room before they buy. The materials array carries your SKU, texture image, and installation location, so the render matches the product, not a lookalike.
Home builders present finish packages on the buyer’s actual elevation and floor plan. Ready2Build runs its selection experience on this API today.
Stage entire listings programmatically — create a project per listing, choose from supported room types and styles, and generate consistent staged sets at listing-feed scale.
Attach the after-picture to every estimate. A homeowner who has seen the result approves faster — and upgrades more.
Questions
A REST API that renders photorealistic renovations onto real photos. Send a photo of a room or exterior plus what to change — free-text guidance or a structured materials array — and receive a rendered image back, typically in 6–15 seconds.
Yes. The materials array carries your SKU, product name, texture image URL, installation location (floors, cabinets, countertops, backsplash, walls, exterior siding, ceiling), and metadata like finish and dimensions. Renders apply your product, not a generic lookalike.
The renders on this page took 6–10 seconds each, measured as full API round-trips. Plan for up to ~20 seconds on large images.
It is a standard REST API with JSON in and a hosted image URL out — use it from any stack. Teams typically ship a product-page visualizer, an internal staging pipeline, or an embedded widget. We work with you directly on integration; most teams have a first render on their own catalog the same day they get a key.
Yes — dedicated endpoints create staging projects and generate styled variations per room, with a catalog of supported room types and staging styles.
API access is provisioned per team. Book a demo and we will set up your key, credits, and a working example against your own catalog.
API access is provisioned per team — we set up your key, credits, and a first working render against your own catalog. Most teams see their product in a customer photo the same day.