JSFiddle VS Decato

Decato's answer:

It starts from the floor plan rather than from a photo or a prompt. The room exists as wall lengths, door swings and clearances before it is ever an image, so every piece is checked against those numbers and anything that will not clear a walkway or a door is dropped before you see it.

What comes back is a specification of real, in-stock products with prices and dimensions, plus a render assembled from those exact items rather than from invented furniture. Most tools in this category end at a picture and leave the sourcing to you; Decato ends at the list you can order from.

Decato's answer:

Because of what you are holding when it finishes.

AI render tools give you an image. You still have to work out what the sofa is, whether it is in stock, what it costs and whether it fits through the door. Traditional planners give you a drawing, and the sourcing is still yours to do. Decato returns the drawing, the products, the fit checks and the client deck in one pass, and every row in the specification links back to the product page.

It is also worth knowing where it does not compete. It is a Mac app; there is no mobile version and Windows is not out yet. Planning is unlimited and free on your own ChatGPT, Claude or Gemini subscription, so you only pay when you want sourcing and photoreal renders.

Decato's answer:

Two groups, doing the same job at different scales.

Interior designers furnishing real client projects. They work to a plan, a budget and a deadline, and what they have to hand over is not a mood board but a specification a client can order from. They are the reason the fit checks and the exported deck exist.

And people furnishing their own flat who have measurements to work from — a room they have just moved into, a renovation, a rental they want to plan before buying anything. Same problem, one room at a time, and usually the first time anyone has told them a 209 cm bed takes almost a third of a 10 square metre bedroom.

Both are on macOS, since that is the only platform Decato ships on today.

Decato's answer:

Decato is a second attempt. The first product was a different Mac app that ran for two years and was stopped for lack of traction — the lesson being that a tool has to finish a job someone is already doing badly, rather than add a step to it.

So this one started with about two dozen conversations with working interior designers. The complaint was consistent, and it was not about renders: making the picture is the easy part now. The hours go into what comes after — identifying every item in it, checking it is still in stock, pricing it, and finding out that the wardrobe you drew does not clear the bedroom door.

Decato was built to end at that point instead of starting there. The floor plan comes first, the geometry is real, and the last thing it produces is a priced list rather than an image. It is early: the Mac app is out, Windows is planned, and there is a lot still to build.

Decato's answer:

• Electron on Node, shipped as a native macOS desktop app rather than a browser tab. Signed and notarised builds for Apple Silicon and Intel, macOS 12 or later, with Squirrel handling updates in the background.
• A geometry engine working in real world units. Walls, door swings, wall contact, front clearance and walkway width are computed against the drawing, not inferred from an image. This is what lets the app reject a placement instead of rendering it.
• Model Context Protocol (MCP) for tool access, which is how the planning agent reaches the drawing, the product catalogue and the render service.
• The user's own AI subscription. The agent runs on your ChatGPT, Claude or Gemini account rather than on credits we resell, which is why unlimited planning can be free.
• Live retailer catalogue data for the product matching, so availability and prices come from the shops rather than from a static library.
• Documents are stored locally as .dec files; presentations export to PDF with the product links intact.