CNC quotation software: how a price is actually built

• 1
  What CNC quotation software reads from your file
• 2
  How the engine picks a machine and a setup count
• 3
  Where the algorithm stops being reliable
• 4
  Using CNC quotation software without getting burned
• 5
  Why shops with real capacity still quote by hand
• 6
  Automated quote vs engineer-reviewed quote

Upload a STEP file and the software starts by reading geometry. It counts faces, finds pockets and holes, measures the smallest internal radius, and works out how much material sits above the finished surface. That stock-removal volume is the raw number behind every cycle-time estimate.

Next it matches features to a tool library. A Ø6 mm flat end mill can clear a pocket, but a Ø2 mm corner needs a smaller tool and slower feed, so the engine adds time. Deep pockets with a 4:1 depth-to-diameter ratio get flagged, because long tools deflect and the software has to slow down to hold tolerance.

Material comes from the dropdown you pick. The engine pulls a hardness and machinability factor, then scales cutting speed and feed. Aluminum 6061 cuts fast and cheap. Inconel does not. The same pocket that takes 8 minutes in 6061 can run past 40 minutes in a nickel alloy, and a quotation engine that only knows the material name will get that spread roughly right, not exactly right.

Tolerance and finish are the inputs most files leave out. If the drawing says ±0.005 mm and Ra 0.8–1.6 μm but the model carries no GD&T, the software assumes a default. That default is usually looser than what the part needs. The quote looks low. The first article then fails inspection, and the real cost lands on the next revision.

Once cycle time is estimated, the software assigns a machine class. A part that fits in a 500 × 500 × 450 mm envelope on three axes usually stays on a 3-axis mill. Add an undercut, a compound angle, or five faces in one setup and the engine routes it to a 4-axis or a simultaneous 5-axis center.

Machine rate is where quotes diverge between suppliers. A 5-axis center carries a higher hourly rate than a 3-axis mill, so the same part can price 30–60% apart depending on routing. Good software shows the routing assumption instead of hiding it behind one number.

Setup count is the second lever. One setup on a 5-axis machine often beats three setups on a 3-axis machine, even at a higher hourly rate, because each extra setup adds fixturing, re-datuming, and queue time. The engine usually gets this right when the part is a single solid. It gets it wrong when the part needs a soft jaw, a custom fixture, or a rotary table.

Batch size changes the math again. Fixture cost and programming spread across the run. At one piece the fixture can dominate the price. At 10,000 pieces the same fixture is nearly free per part, and cycle time takes over. Software that prices every quantity off one setup assumption will mislead you at one end of that range.

Automated quoting works best on prismatic parts: brackets, housings, plates, manifolds, turned shafts with clear features. These have geometry the engine can decompose into tools and passes. Quote accuracy on this family is often within 10–15% of a manual quote.

It degrades on thin walls. A 0.8 mm wall in aluminum will chatter unless the programmer adds support, reduces radial engagement, and may need a different workholding plan. None of that is in the feature tree. The engine sees a pocket. The machinist sees a part that will move.

It also struggles with tolerance stacking and inspection cost. A part with twelve bores at ±0.01 mm and a true position callout needs a CMM program, a fixture, and time on the metrology bench. Software that prices only machine time will quote the cutting and forget the checking.

Surface finish follows the same pattern. Bead blasting, anodizing, and laser marking are line items the engine can add. Hardcoat anodizing at a controlled thickness, or a medical finish that needs passivation documentation, is a conversation. Ask for the manual review whenever the finish spec runs longer than one line.

Treat the first number as a screening tool, not a contract price. It tells you whether a design is in the ballpark of your budget. It does not tell you what the part costs to make well.

Attach the drawing, not just the model. GD&T, finish callouts, material certification requirements, and inspection level change the price more than most geometry edits do. A quote built from a bare STEP file is an estimate with a wide band.

Ask what the rate includes. Programming, fixturing, first-article inspection, and material certification are sometimes in the hourly rate and sometimes billed separately. Two quotes at the same number can mean different scopes.

Send the same part to two suppliers and compare routing, not just price. If one quote assumes 5-axis in one setup and the other assumes three 3-axis setups, the difference is real and the price gap is explainable. That comparison teaches you more about your part than any single number.

For anything with a safety function, a regulated industry, or a tolerance tighter than ±0.01 mm, skip the instant path. A DFM review catches the features that will fail before the tool is cut. At GreatLight, that review comes back with the quote inside 12 hours.

A shop running 127 high-precision CNC machines, including 16 simultaneous 5-axis centers, cannot let software set the price alone. Machine scheduling, tool availability, and operator skill all move the real cost, and none of them live in a file.

The working model is hybrid. Software generates the baseline: cycle time, material cost, machine class. A manufacturing engineer then adjusts for fixturing, inspection, and risk. The customer sees one number, but that number passed a human.

This matters most on first articles. A part that has never been machined carries unknown risk: workholding, tool reach, chip evacuation in a deep cavity. A pure algorithm prices the drawing. An engineer prices the drawing plus the things that go wrong on the first run.

That is also why quoting speed and quoting accuracy are not the same claim. A fast quote is a service. An accurate quote is a judgment. When both come from the same supplier, you get a usable number instead of a starting bid.