A founder with a large plastic housing gets two quotes. One shop wants $45,000 for an injection mold and quotes $6 per part. The other wants $4,500 for a thermoform tool and quotes $28 per part. Both are honest quotes for the same shape, and which one is right depends entirely on how many units get built and how the part has to perform. Choosing wrong costs either a year of unnecessary tooling spend or a per-part price that makes the business model impossible at scale.
What each process actually does
In thermoforming — vacuum forming is the most common variant — a sheet of plastic is heated until it is pliable, then pulled down over or into a single-sided tool by vacuum. The sheet takes the tool's shape, cools, and is trimmed. Only one side of the part touches the mold, so only that side is precisely controlled. Pressure forming is the upgraded version: it adds compressed air on the back side, typically 40–60 psi, which pushes the sheet into much finer detail and produces sharp corners, crisp text, and molded-in features that plain vacuum forming cannot hold. Our overview of vacuum forming and thermoforming for low-volume parts walks through the variants.
In injection molding, molten resin is forced under high pressure into a closed two-sided steel or aluminum cavity. Both surfaces are formed, wall thickness is controlled everywhere, and features such as ribs, bosses, snap fits, and living hinges are molded in the same shot. The tool is far more expensive, far slower to build, and produces a finished part in seconds.
The numbers that decide it
| Thermoforming | Injection molding | |
|---|---|---|
| Tooling cost | $1,500–$15,000 typical | $12,000–$80,000+ typical |
| Tooling lead time | 2–5 weeks | 8–16 weeks |
| Cycle time | 30 seconds to several minutes | 15–60 seconds, often multi-cavity |
| Practical part size | Very large — 4 ft x 8 ft panels are routine | Limited by press tonnage and shot size |
| Wall thickness control | Thins as the sheet stretches, often 30–50% in deep corners | Controlled by design across the whole part |
| Tolerances | Roughly ±0.030 in on formed features | Roughly ±0.005 in achievable |
| Molded-in features | Limited; bosses and undercuts usually added later | Ribs, bosses, snaps, hinges all in one shot |
| Secondary work | Always — trimming, drilling, often CNC routing | Minimal, sometimes only degating |
The crossover in real projects usually falls somewhere between 1,000 and 5,000 parts per year for a mid-size housing, but that range moves a lot with part size. A large enclosure that would need a 1,000-ton press pushes the crossover much higher, because the injection tool is enormous. A small part with fine features pushes it lower, because thermoforming cannot make it at all. The general framework for this class of decision is in how to choose a manufacturing process by volume.
Where thermoforming clearly wins
- Large, relatively simple shells. Medical cart covers, kiosk panels, machine guards, spa surrounds, vehicle interior panels, trays and packaging.
- Low to moderate volume. A few hundred to a few thousand units a year, especially for capital equipment sold in small numbers.
- Programs that need to launch fast. A tool in three weeks instead of three months can be worth more than the per-part difference, particularly when the alternative is missing a trade show. The realities behind that gap are covered in how long it takes to build an injection mold.
- Designs likely to change. Thermoform tools are often machined aluminum or even filled epoxy, so a revision costs thousands rather than tens of thousands.
- Products needing many variants. Several tools at thermoform prices can cost less than one injection tool with interchangeable inserts.
Where injection molding wins
- Volume. Past a few thousand parts a year the per-part gap compounds until the tooling looks cheap. Run the payback arithmetic explicitly using the breakdown in injection molding costs.
- Integrated function. If the part needs snap fits, screw bosses, sealing lips, and stiffening ribs, molding gives them free where thermoforming means bonded-on hardware and labor.
- Dimensional precision. Anything that must locate a PCB, seal against a gasket, or mate to another molded part.
- Both surfaces cosmetic. A thermoformed part has one controlled face; the other shows stretch marks and texture from the sheet.
- Thin, consistent walls. Molding lets you specify the wall, which drives strength, weight, and cost — see wall thickness for injection molded parts.
Designing for the process you chose
The two processes want genuinely different geometry, and porting a design across without a redraw is where projects go wrong. Thermoformed parts want generous draft — 3° to 5° on female tools, more than injection molding needs — plus large radii, shallow draws relative to their width, and no undercuts unless the tool has moving sections. Draw ratio is the number to watch: as depth approaches the opening width, the sheet thins badly at corners, so you compensate by starting with heavier sheet, which costs material on every part. Both processes share the same underlying logic on release, described in draft angles and parting lines.
Materials also differ. Thermoforming runs extruded sheet: ABS, HIPS, PETG, polycarbonate, acrylic, HDPE, and ABS/acrylic cap-sheet laminates that give a UV-stable colored surface with no painting. Injection molding runs pellets and has a far wider palette, including glass-filled and flame-retardant grades that are hard to source as sheet.
Two more options worth knowing
For hollow, closed shapes — tanks, bins, large ducting — rotational molding beats both, with cheap tooling and uniform walls at slow cycle times. And twin-sheet thermoforming forms two sheets and fuses them, producing a hollow double-walled part with far more rigidity than a single shell, which is how many pallets, cases, and equipment doors are made.
A pattern worth copying: launch on thermoforming, switch to injection molding when the volume arrives. Design the part so the transition is possible — keep wall thickness realistic for both, keep features that would be molded in as separate bonded parts for now, and freeze the outside surface so the product does not change appearance mid-life. Many capital equipment makers never switch at all, because their annual volumes never justify it.
Projects House takes housings from concept through the process decision, tooling design, and first-article approval, including the cost model that says which route pays back at your real volume. Send us the part and the annual numbers through our contact form.