How much does a 3D configurator cost?
There is no single price: a 3D configurator depends on the number of products and variants, the state of the 3D models, the complexity of the rules, the technology chosen and the integration with your ERP or store. Two projects with the same product count can cost very different amounts.
Why there is no single price
A configurator is not a product sold by the metre. It is custom software that translates a factory’s rules into an interface. The budget measures 3D work, logic work and integration work, and all three change from one manufacturer to the next.
Two manufacturers with the same number of references can get very different quotes. One arrives with clean CAD, written rules and a documented API. The other arrives with PDF drawings, rules that live only in the product manager’s head and a 2009 ERP with no documentation. The second project is not more expensive out of spite: it is more expensive because what is missing has to be built.
That is why we distrust headline figures. A low range usually describes a catalogue product with fixed options. A configurator with manufacturing rules, configuration-based pricing and order integration is different work. Asking for the price before defining the scope is asking for an answer nobody honest can give.
How many products and how many variants
The first factor is the number of products, and above all the number of combinations per product. Five products with ten options each is one project. Five products with free sizes, six fabrics and modules that combine is a very different one. Combinatorics drive the budget.
Every option adds rules: what pairs with what, which size range is valid, which surcharge applies. Ten independent options are a thousand possible combinations; if three of them condition each other, the number of rules grows faster than the number of options.
The number of models to prepare counts too. A product with three finishes needs one optimised model and three materials. A product with twelve structural configurations may need twelve meshes or an interchangeable-parts system. The second is more 3D engineering and less content work.
Where the geometry comes from
The state of the 3D is the biggest cost driver and the one least visible in a demo. A SolidWorks model ready to optimise does not cost the same as rebuilding a product from drawings or photographs. The difference can be half the project.
One detail gets forgotten: a render model and a configurator model are not the same object. The render model can carry millions of polygons, baked lighting and welded parts. The configurator model needs separate parts, stable material names and a weight a web page can carry. Turning one into the other is work, even when the file already exists.
| Starting point | What has to happen | Effect on the budget |
|---|---|---|
| Clean CAD (STEP, IGES, SolidWorks, Rhino) | Optimise the mesh, separate parts, apply materials | The most contained scenario |
| Existing render model | Review topology, adapt weights and materials | Moderate, if the geometry holds up |
| Drawings and dimensions | Model from scratch at web-level detail | High: it is engineering, not conversion |
| Photos or one physical unit only | Model and validate against the real product | Highest: there is back and forth |
Configuration and pricing rules
Rules are the part no template solves. The closer your catalogue is to a matrix of exceptions — non-standard sizes, modules that do not combine, surcharges per finish — the more work goes into writing, testing and documenting them. This is where a project is won or lost.
A badly written rule becomes an impossible order. That is why the budget includes time with your team: sitting down with product and with the factory, listing forbidden combinations, deciding what happens when a buyer goes out of range, and checking that the price on screen matches the price the ERP calculates.
If price is fixed per reference, this part is short. If price is calculated — per linear metre, per module, per fabric, per machining hours — you need a calculation engine validated against real cases. That is where the difference between what sales quotes and what the factory invoices shows up.
The technology you choose
The three technologies differ in build cost and in running cost. Pre-rendered images move the cost to the start — every visible combination has to be rendered — and then behave as files. WebGL concentrates the work in optimisation and interface. Pixel streaming adds a cloud GPU per user.
The choice is not only technical. A catalogue with 400 combinations can be solved with images and selection logic; trying to render all of it in real time multiplies optimisation work without improving the buying decision. The opposite is also true: a product the buyer wants to open and turn needs WebGL even with fewer combinations.
| Technology | Build cost | Running cost |
|---|---|---|
| Pre-rendered images | High upfront: one view per visible combination | Low: they are files on a CDN |
| Real time (WebGL) | Medium-high: optimisation, materials and interface | Low: the work happens on the device |
| Pixel streaming | High: engine scene, video and session handling | Per connected user: one cloud GPU |
Integration with your system
Connecting the configuration to an ERP, CPQ or store is a project inside the project. A well-defined JSON against a documented API is contained. Writing into an old ERP with a proprietary format and no test environment can cost more than the whole interface.
The budget should separate three things: the configuration, the output in an agreed format, and the connection to the destination system. That way you can launch with a JSON output and connect later without blocking the release. It is also worth knowing who maintains that connection when your ERP is updated.
With a CPQ, integration is usually cleaner because the system already expects configurations. With Excel and email, part of the budget goes into defining a format that does not exist yet.
Design, content and accessibility
The interface has a real cost: it is not a form with a viewer next to it. Someone has to decide the order of decisions, what shows on mobile, how a surcharge is explained and what happens when a combination is unavailable. Someone has to write the copy and check accessibility.
A badly ordered configurator makes the buyer pick the fabric before the size and discover at the end that their combination does not exist. Ordering the decisions is product design work, and it shows in conversion more than the realism of the render.
Accessibility is part of the budget too: keyboard navigation, visible focus, alt text, contrast and 44 px touch targets. It is not an extra added at the end; left for review, it forces components to be rebuilt.
The invisible cost: keeping the catalogue alive
A configurator keeps living after delivery. New fabrics arrive, prices change, finishes are retired and products are added. That maintenance is a recurring cost that never shows on the first invoice, and it is worth deciding from the start who does it, how often, and with which tools.
With custom development the code and models are yours, so you can maintain them with your team or another supplier. There is no platform fee. But there is work: every season brings changes, and if the rules live in the code instead of a panel, every change needs someone who knows how to touch it.
Part of the budget should go into making that maintenance cheap: rules separated from the interface, stable material names, documentation and a simple panel to edit options and prices. That is the difference between an asset and a dependency.
How to ask for a quote you can compare
Ask for scope, not a loose figure. A comparable quote lists products, variants, rules, technologies, integrations, deliverables and what is excluded. With that list you can put two proposals side by side and see what each one includes, instead of comparing two numbers that measure different things.
- A closed list of products and variants in the first phase.
- State of the geometry: which models you deliver and in what format.
- Combination and pricing rules, with the edge cases identified.
- Proposed technology and why, with the performance target in writing.
- Integration: which system, which output format, and who maintains it.
- Deliverables: code, models, documentation and an editing panel.
- What is out of scope, and how it would be quoted later.
- A phased timeline with milestones you can test.
Frequently asked questions
How much does a 3D configurator cost?
It depends on the number of products and variants, the state of the 3D models, the complexity of the manufacturing and pricing rules, the technology and the integration with your ERP or store. We do not give a figure without knowing the scope: after a 30-minute call we send a closed proposal.
Why can a cheap configurator end up expensive?
Because the initial invoice does not include what comes later: rules the template cannot handle, integrations that have to be built separately, changes that depend on the supplier and a fee that grows with the catalogue. Comparing only the first bill hides the three-year cost.
Is there a monthly fee?
With custom development, no. There is no platform in between: you pay for the project and the code and models stay in your hands. There are hosting and maintenance costs if you want someone to handle catalogue updates.
What can I do to keep the cost down?
Arrive with clean CAD, a closed list of options and the rules written down. Start with one product or one family and expand later. And settle the integration level early: changing it mid-project is what makes it most expensive.
Can we start with a single product?
Yes, and it is usually the sensible choice. One well-resolved product validates the rules, the performance and the order flow before you invest in the full catalogue. Families are added afterwards on the same structure.
Is maintenance included?
Delivery includes documentation and a panel to edit options and prices, so your team can handle day-to-day changes. If you prefer us to do it, that is a separate agreement with a concrete scope.
How long does it take?
A product with material variants usually runs four to eight weeks. A wide catalogue with manufacturing rules and ERP integration runs ten to fourteen. The calendar depends mostly on the availability of your data and your rules.
