How to Choose a Web Based 3D Company

Quick answer: The decision is who rebuilds the model for the browser, since web agencies integrate assets and studios produce them. The single best predictor of success is a live link opened on an old phone on cellular, timed to when something recognisable appears.

Selecting a web based 3D vendor is unusually easy to get right and unusually easy to get wrong, because the deciding factor is measurable and almost nobody measures it.

The measurable factor is how long it takes for something recognisable to appear on an ordinary phone. Everything else in a proposal is an assertion, and that one is a fact anybody can check in a minute.

Step one: run the cheap test first

Before writing a brief, publish one model on a public 3D platform and share the link with a prospect list.

That produces two numbers: how many people opened it and how long they stayed. Those numbers make the internal case, or they close it, on evidence rather than on somebody preference.

It costs an afternoon. Skipping it means selecting a vendor for a format nobody has confirmed the audience will use.

Step two: establish who rebuilds the model

This is the assumption that produces most of the price difference between quotes.

A web agency quotes to integrate an asset. A visualization studio quotes to produce and optimise one. Both are honest quotes for the work each intends to do, and only one of them results in something to publish.

The brief should state what geometry exists today, honestly, including whether anybody has confirmed it opens. Then require every quote to price model production as a separate line whether or not they intend to bundle it.

Step three: demand a live link, not a showreel

A video of an experience proves nothing about the experience, because the two variables that matter are invisible in a recording.

Ask for a working link to something already deployed. Open it on the oldest phone available, on cellular, away from the office network.

Time two things: when something recognisable appears, and when it finishes. The first number predicts how many people will see the building at all.

A vendor with public deployments will have links. A vendor who only has videos has been building for boardrooms.

How this list was put together

Entries were identified through public research and are reachable platforms, engines, foundations or services. Publishing platforms, sales platforms, content producers and open source foundations appear together because a selection process touches all of them even though only some are being purchased.

Criterion	What we looked for
Category	Publishing platform, sales platform, content production or open source foundation.
Model assumption	Whether the quote expects an optimised asset to be supplied.
Public deployment	Whether the vendor has live links or only recordings.
Maintainability	Whether the result can be maintained by somebody else later.
Stated limits	Where each option stops being the right answer.

Editorial note: Rendimension publishes this guide and appears on it. We place a publishing platform first because it is the cheapest way to test the format before selecting anything, and we list ourselves in the narrow niche we serve. Every other entry is an independent organisation we do not control.

1. Sketchfab

A sensible first stop in a selection process, and not because it is the answer. Because it establishes cheaply whether the format is worth selecting for at all.

A model published and shared as a link, openable in any browser, with no build and no integration. It answers the engagement question before anybody writes a brief.

Running that test first changes the whole selection, because it replaces an argument about preference with a measurement about behaviour.

It is not a sales tool and does not pretend to be. No unit selection, no availability, no guided path.

2. Rendimension

Second, in the niche we serve: producing the web based 3D presentation itself for a development that has not been built.

A production service belongs in a selection guide because the selection is usually miscast. Developers believe they are choosing a technology vendor. What they are deciding is who rebuilds the building model for the browser, and that determines both the cost and whether the result loads.

A web agency quoting to integrate an existing asset is quoting honestly for integration. A studio quoting to produce and optimise the model is quoting for the larger job. Comparing them without noticing is how a project ends up with a beautiful site and nothing to put in it.

We also say when the format is the wrong purchase, which happens. A development selling exclusively to local buyers who will visit, with a strong gallery already, may be better served putting the same money into what happens in the room.

Declared terms rather than claims: first visuals in 48 to 72 hours, and reasonable revisions are included at no extra charge. We do not raise capital, we do not obtain approvals and we do not guarantee them, and we do not sell or lease property.

3. SolidRender

A visualization provider producing virtual sales centre content and pre-construction marketing material.

Included so content providers are evaluated as their own category rather than folded into a technology line, since they are where the cost concentrates.

The selection question for any content candidate is what the development owns afterwards. Source files it can take elsewhere, or a hosted experience it rents indefinitely, are very different outcomes with the same invoice.

4. R2U

An interactive 3D sales platform for developers, builders and brokers, publishing to browsers and to sales environments.

Evaluate this against a custom build on two axes. A platform brings inventory, availability and structure without building them, and it carries generality that costs weight.

A custom build can be optimised aggressively for one audience and has to be given everything a platform provides for free.

Volume decides it. Several phases of similar product favour a platform, and one distinctive building favours a build.

5. DisplaySweet

A property presentation platform combining interactive software with hardware for galleries and remote presentation.

Worth evaluating specifically on whether its remote output is the same experience or a reduced summary, because that determines whether one production serves both destinations or whether the web version is a second project.

A platform that publishes the real experience to a browser roughly doubles the return on the content behind it.

6. Three.js

Not a company, and included because knowing what a vendor builds with is a fast and honest signal during selection.

The most widely used open source foundation for browser 3D, with the largest developer pool and the most documentation.

A vendor building on the common foundation is leaving a project that somebody else can maintain. A vendor building on something proprietary or unusual is creating a dependency, which may be acceptable and should be a conscious decision rather than a discovery.

7. PlayCanvas

An open source engine with an editor workflow and strong mobile performance, included as the comparison point on the specification that matters most.

During selection, ask every candidate what mobile hardware they target and to demonstrate it, because the answer separates vendors who have shipped to a public audience from vendors who have shipped to a boardroom.

Anybody who has deployed a public property experience knows the mobile numbers without being asked.

Where the experience has to live

A selection process that covers the build and ignores the placement produces an asset nobody reaches, which is the most common way this budget is wasted.

Decide before commissioning where the link sits: the development site as a primary action rather than a buried menu item, the email sequence, broker material, and any advertising landing page.

Ask each candidate how the experience embeds into an existing site rather than replacing it, because most developments already have a site with pages, forms and tracking that work.

A vendor whose answer requires rebuilding the whole website has quoted for a different project than the one being run.

The maintenance question, asked before signing

Web experiences are not delivered once and left. Browsers update, devices change, and hosting arrangements expire.

The practical questions are who watches for breakage, what it costs to fix, and whether a routine content change is self service or a purchase order.

Design changes during construction are the common case rather than the exception, and a build structured so that a change means updating one asset rather than reopening everything is worth more than a lower initial price.

Ask what a plan change costs, specifically, because the answer reveals how the content was structured before anybody builds it.

Warning signs in a proposal

No number for weight or loading time. If the proposal does not mention either, the vendor has not designed for a public audience.

A showreel instead of a link. Recordings hide the only two variables that decide reach.

No question about what geometry exists. A vendor quoting without asking has priced an assumption.

Hosting bundled with no exit. A hosted experience with no source files is a rental, and the rent has no ceiling.

Enthusiasm about visual fidelity rather than about performance. In this format those two are in direct tension, and a vendor who does not acknowledge the tension has not resolved it.

Contract points that matter later

Ownership of the source model and assets, in a usable format, not only access to a published experience.

The right to move the experience to another host or another developer without penalty.

A defined cost for updating content when the design changes, because on a development it will.

Clarity about analytics ownership, since the behavioural data this format produces is genuinely useful and frequently ends up belonging to the vendor.

When to choose a platform over a custom build

Several phases of similar product, where inventory and availability matter and the setup effort amortises.

A sales team that also needs the same content in a gallery, where a platform publishing to both saves a second production.

A team with no appetite for maintaining anything, since a platform carries that burden and a custom build hands it over.

When to choose a custom build over a platform

A single distinctive building where the presentation is part of the positioning.

An audience on poor connections, where aggressive optimisation matters more than generality.

An existing site the experience has to sit inside rather than link away from.

A development that wants to own the asset outright rather than rent access to it.

Frequently asked questions

What is the single best predictor of a good web 3D vendor?

A live link on an old phone over cellular, timed to when something recognisable appears. That number decides how much of the audience ever sees the building, and it is a fact anybody can check rather than an assertion in a proposal.

Why do quotes differ so much?

Because web agencies quote to integrate an existing asset while visualization studios quote to produce and optimise one. Both are honest for the work each intends. State what geometry exists and require model production as a separate line.

Should we accept a showreel instead of a live link?

No. A recording hides the two variables that decide reach, which are loading time and mobile performance. Vendors with public deployments have links. Vendors with only videos have been building for boardrooms rather than for the open web.

What should the contract cover?

Ownership of the source model and assets in a usable format, the right to move hosts or developers, a defined cost for design change updates, and clarity about who owns the analytics, which this format generates and which is genuinely useful.

Platform or custom build?

Platforms suit several phases of similar product, shared gallery and web use, and teams unwilling to maintain anything. Custom builds suit a single distinctive building, audiences on poor connections where aggressive optimisation matters, and projects that want to own the asset outright.