What is the difference between a looks-like and a works-like prototype?
According to the University of Maine Advanced Manufacturing Center, a "looks like" prototype demonstrates color, shape, feel and size. It is built to be seen and handled, not stressed or operated. The source notes that "many rapid prototype or 3D printed parts fall under this category," which is a big reason 3D printing has become the default first step for so many inventors.
A "works like" prototype is a different animal. The same source calls these "proof of concepts" and says plainly that "they may not be very pretty, but they will prove that an idea works." The job of a works-like prototype is function, not appearance. A rough-looking box of wires and gears that actually does the job counts as a works-like prototype. A beautiful 3D-printed shell that cannot survive being picked up does not.
The University of Maine also describes a later stage that combines both: "looks like and works like" prototypes, which it places later in development, closer to something you would show a manufacturer or investor as a near-final version.
What is a 3D print actually good for?
A 3D print is a strong tool for the looks-like stage. It lets you hold your invention, check proportions, test how it feels in the hand, and see whether parts fit together the way you imagined on paper. This is also where fit checks live: does the lid snap onto the base, does the part slide into the housing, does the overall size feel right next to the product it is meant to work with.
If you have not yet turned your idea into a digital model, that step usually comes first. EntreDash's invention 3D model service builds a print-ready STL, OBJ and GLB file from $39.99, and the sketch to 3D model guide walks through getting from a drawing to something you can print.
Once you have a model, our pillar guide on 3D printing a prototype covers the process end to end, and the FDM vs SLA vs SLS comparison and where to get a prototype 3D printed posts go deeper on choosing a process and a provider.
For straightforward looks-like parts, Fictiv's prototyping guide notes that simple 3D-printed parts can run under $50 [guide]. That is often enough to answer the basic question: does this look and feel right.
When is a 3D print not enough?
A 3D print struggles once your invention needs to prove function rather than form. If your idea involves moving parts under load, a living hinge that has to flex thousands of times, a part that needs to seal against water or pressure, or a material property like flexibility or heat resistance, a plastic 3D print often cannot answer those questions honestly. A print that looks finished can quietly mislead you into thinking the design works when it has not actually been tested that way.
This is the point where many inventors move to CNC machining, urethane casting, or bring in a design engineer or product design firm.
What does CNC machining add?
CNC (machined) parts let you test with materials closer to what a final product might use, like aluminum or engineering plastics, and they hold tighter tolerances than most desktop 3D printing. Protolabs says CNC machined part prices start around $65, with lead times of 1 to 3 business days once you upload a CAD file. Fictiv's guide places CNC prototypes more broadly at about $100 to $1,000 or more, depending on complexity [guide].
What does urethane casting add?
Urethane casting is often used when you need a handful of parts in a rubber-like or plastic-like material that behaves closer to an eventual injection-molded part, without paying for full production tooling. Fictiv's guide puts urethane-cast prototypes at roughly $200 to $5,000 or more [guide].
What does a design firm or engineer add?
A design engineer or product design studio can help when the question is not just "can this be made" but "will this actually work as intended." They can build and test a true works-like prototype, sometimes well before anything looks finished. One studio, Analogy Design, publishes its own pricing as a reference point: looks-like prototypes at $300 to $2,000, works-like at $2,000 to $20,000, alpha builds at $15,000 to $40,000, and production-intent prototypes at $30,000 to $150,000 or more, with connected hardware (sensors, electronics, apps) adding another $5,000 to $30,000 [guide]. These are one firm's published numbers, not a universal price list, but they give a sense of how much the cost grows as a prototype moves from appearance to function to near-production.
If you are trying to get a rough sense of your own project's likely costs across these stages, EntreDash's hardware development cost calculator is free to use and can help you think through where your invention is likely to land.
How do I decide which one I need right now?
A useful way to think about it: ask what question you are actually trying to answer. If the question is "does this look and fit the way I pictured it," a 3D print is probably enough, and it is the cheaper, faster path. If the question is "does this actually work when someone uses it the way it is meant to be used," you likely need a works-like prototype, which may mean CNC, casting, or outside engineering help.
It is also fine, and common, to do both in sequence. Many inventors start with a looks-like 3D print to nail down size and shape, then move to a works-like version once they know the shape is right and are ready to test function. The University of Maine's "looks like and works like" category describes the combined version that comes later, once both questions have been answered separately.
What comes after the prototype is built?
Once you have a prototype, whether it is a looks-like model or a working version, the next step is usually talking to a manufacturer. That conversation goes a lot smoother when you can show up with a clear model or sample and explain what you need. EntreDash's guide on explaining your invention to a manufacturer covers how to present your idea clearly, what questions to expect, and how to avoid common misunderstandings at this stage.
Before sending files or samples widely, it is worth pausing to think about what you want to protect and how. An NDA or a provider's confidentiality terms are not the same thing as patent protection, and they do not make an idea patentable on their own. If you are unsure what to do before sharing your design broadly, EntreDash's post on NDAs before sharing an invention is a good place to start, and speaking with a patent attorney or agent before sharing files widely is generally a smart move.
What is the short version?
A looks-like 3D print answers questions about appearance, size and fit, and for many early-stage inventions that is exactly what you need first. A works-like prototype answers whether the thing actually functions, and that often calls for CNC machining, urethane casting, or help from a design engineer. Matching the prototype type to the question you are trying to answer will save you money and keep you from drawing the wrong conclusions from a model that was never built to be tested that way.


