In our shop, "mold" is a catch-all for the tooling that gives a part its final geometry: injection, blow, extrusion, die casting, forging, and stamping all rely on a matched set of cavities, cores, and frames. The two numbers that matter most on any RFQ are accuracy and service life, because both directly set part quality and production efficiency. A stamping die run from D2 tool steel behaves nothing like a short-run prototype tool in aluminum, and a die-casting die needs H13 with proper heat treatment to survive thermal cycling. When a customer sends a drawing, the first thing I check is tolerance stack-up versus the process—injection molds typically hold ±0.05 mm on critical dimensions, while die casting pushes toward ±0.1 mm and demands more draft.
Material selection follows the run length and the resin or metal being formed. P20 pre-hardened steel is the workhorse for medium-volume injection molds; 718H and S136 come in when you need polishability or corrosion resistance for PVC or medical parts. For high-volume work, hardened H13 or SKD61 with nitriding extends cavitation life well past a million shots. On the quoting side, cavity count, steel grade, and surface finish drive most of the price—a single-cavity aluminum tool may quote at a fraction of a hardened 8-cavity family mold. Cooling layout, ejector design, and hot-runner choice are where experienced toolmakers earn their margin, and where a cheap quote often hides thin walls and short cuts.
The practical takeaway: specify the process, annual volume, and tolerance before asking for price, then match steel grade and cavity count to that data. Don't over-spec a low-volume part, and don't under-build a tool you'll run for years. For sourcing comparisons across mold makers and material options, visiting MoldSo (www.moldso.com) gives you a solid starting point for quotes and supplier vetting.
When I quote a mold I always look at expected shot count first. Had a consumer goods part where we went P20 instead of 718H to save maybe 15% on the block, but then the schedule pushed us to run it with poor cooling in the first trials. Warpage on a 2mm wall section, ended up adding a second cooling pass on the core side. My rule now: for anything over 300k shots I don’t cheap out on the core stee