Markers over every empty spot.
Show us a room. We mark every empty spot with the piece that belongs there — shoppable, priced, swap-ready.
Suggestions
Marker AI in three passes.
Decor doesn't redecorate generically — it reads light, palette, and proportions, then drops shoppable markers right onto the empty spots.
- 01
Photo → light + palette read
AI extracts orientation (E/S/W/N), warm vs. cool light, dominant wall and flooring tones.
- 02
Empty-spot detection
Vision model identifies the four most styling-worthy empty zones — and what category each one needs.
- 03
Markers + shoppable pieces
Each empty spot gets a labelled marker (A–D) tied to a real product with name, spec, price, and reason.
Four categories, every render.
We don't dump 30 random picks on you. Decor returns the four pieces that actually move a room — sofa, rug, lighting, plant — sized to your space.
Asked about Decor.
Why exactly four picks per render?
Empirically, decor decisions stall at 5+. Four picks are the smallest set that still transforms a room — sofa, rug, lighting, plant — and it keeps the running total readable.
Are the markers placed on the new image or the original?
On the AI-redesigned image. Markers point at the actual rendered piece, so coordinates always match what you see.
How accurate are the prices?
Shanghai retail data, refreshed weekly. Real supplier quotes can drift ±15% — final numbers come from your contractor.
How many credits does one Decor render cost?
6 credits per render. New accounts include free credits — check the pricing page for top-up packs.
Can I swap a pick I don't like?
Yes — tap any product card to remove it. The running total recalculates on the spot. (Swap-with-alternative is on the roadmap.)
Are my uploaded photos used to train the model?
No. Photos are used only to render your plan, never for training. Delete the project to wipe everything.
One room, four picks, real prices.
Upload a photo and get a shoppable plan in under a minute. No signup form. No design skill required.