A press brake’s price is built from five cost blocks: structure, power and drive, control and axes, tooling, and compliance with support. They do not grow at the same rate, and they are not equally visible. Two machines with the same tonnage and the same bending length can be quoted far apart because the cheaper one has moved cost out of blocks you cannot inspect once the frame is painted.
This article does not give you a number, because any figure published online — including any figure we might publish — is a snapshot of one configuration, one steel price, one exchange rate and one freight market on one day. What it gives you is a way to read a quotation, and a way to defend the decision afterwards.

The Five Blocks Behind Every Quote
| Cost block | What it physically is | Grows with | What to ask for in writing |
|---|---|---|---|
| Structure | Welded frame, stress relief, machined bed and ram, uprights | Tonnage, bending length, throat depth | Plate thickness at the critical sections, stress-relief method, machining tolerance on the die seat |
| Power and drive | Cylinders, pump, valves, reservoir — or servo motors with screws or belts | Tonnage, drive architecture, ram speed | Motor power, pump and valve brands and model numbers, whether pressure is held or released between strokes |
| Control and axes | Controller, per-axis drives, position feedback, software options | Axis count far more than tonnage | Controller model and the axis letters it actually supports, feedback device per side, whether offline programming is licensed |
| Tooling and clamping | Punches, dies, holders, crowning mechanism | Tooling length and standard, not tonnage | Tooling standard, material and hardness, maximum load per metre for that specific tool, segment lengths supplied |
| Compliance and support | Guarding, conformity documentation, commissioning, training, spare parts | Destination market and service level | Guarding type and rating, the standard the machine is built to, training days, response commitment, parts lead time |
The blocks do not have the same slope, and this is the part buyers miss.
Structure and power climb steeply with size. More tonnage means thicker plate in the frame, larger cylinder bores, a bigger pump, larger valves, a larger reservoir, heavier machining and heavier handling. Each step up is not proportional to the tonnage number; it is steeper, because several of these items scale together.
Control and tooling barely move with tonnage at all. A control package with six axes, a graphic controller and per-side position feedback costs broadly the same whether it is mounted on a 40-tonne frame or a 300-tonne one: what changes is the size of the axes it drives, not the architecture of the system. The same is largely true of a motorized crowning system and of a set of precision-ground tooling of a given length.
That asymmetry explains most of the pricing behaviour that confuses buyers. On a small machine a high specification is a large share of the invoice, and it is the easiest place for a supplier to move the number in either direction. On a large machine the same specification is a small share, so arguing over it saves less than arguing over structure and drive. It also tells you where to spend: options are cheapest, in relative terms, on the machine you are already buying for its tonnage.

Drive architecture is a step, not a slope
The three families buyers actually compare — mechanical torsion-bar synchronization, electro-hydraulic servo synchronization, and full servo drive using screws or belts — sit in three price bands because each carries a fixed block of content the others do not: a torsion bar and mechanical linkage; proportional valves, per-side feedback and a closed-loop controller; or servo motors and the transmission and frame layout that suit them. You are not paying for a label, you are paying for the components that make the label true. Which architecture suits your parts is a separate question, set out in servo-electric vs hydraulic vs hybrid press brakes and in CNC vs NC vs torsion-bar press brakes.
One pricing detail hides in the safety standard. ISO 6909:2026, published in May 2026, applies to four machine types — hydraulic, hydraulic servo-drive, screw servo-drive and belt-spring servo-drive — and states plainly that it does not cover machines using a combination of technologies (ISO 6909:2026). Two consequences for a buyer: conformity is an obligation on the party that designs, manufactures and supplies the machine, so it is a cost inside the price rather than an extra; and a hybrid machine falls outside that scope, so its conformity route has to be argued on its own terms rather than assumed.
Tonnage and Length: Two Different Cost Curves
Tonnage is not bought linearly. Force comes from a system — cylinder area, hydraulic pressure, pump displacement, valve size, frame stiffness and the tooling that transmits it. Ask for twice the force and you do not get twice the steel: you get a heavier frame in the sections that carry the load, bigger cylinders, a larger hydraulic package, and more machining and handling hours on every one of those parts.
The force itself is easy to check. For 450 MPa mild steel in air bending, required force runs at roughly 65 × S² ÷ V tonnes per metre, with thickness and die opening in millimetres and the V opening at about 8 × S. That is why a 100-tonne machine over 3.2 m lands at around 3.5 to 4 mm of mild steel in most specification tables, and why stainless of the same thickness needs roughly 1.4 times the force. Run your part through the tonnage calculator before you discuss price; a machine bought on nominal tonnage alone is frequently either oversized or quietly incapable.
There is also a limit inside every rating that the nameplate does not show: full tonnage cannot be concentrated on a short window between the uprights. The common working rule is that full force should not be applied over less than roughly 60% of the distance between uprights, and the tooling carries its own load limit per metre, which is usually the binding constraint on short, heavy work.

Length is worse than linear. A 4 m machine does not cost twice a 2 m machine, and the reason is not marketing. Bed and ram are single pieces of steel that must be machined straight over their whole length, stress-relieved, transported and lifted as one unit. Deflection compensation has to work across the full span, which on long machines means several independently adjustable points rather than one. Tooling has to be supplied in longer, straighter, segment-matched lengths. Two 2 m machines are also not a substitute: they need two operators and twice the floor space, which the arithmetic usually hides.
What the 4 m premium actually buys is a single-setup capability — one part, one handling, one operator, angle consistency end to end. If your work is long panels, that is a production capability rather than capacity, and on that work it is worth more than the same money spent on tonnage.
| 170 t / 3,200 mm | 170 t / 4,100 mm | |
|---|---|---|
| Typical work | Medium panels, enclosures | Long panels, full-length profiles |
| Where the extra money goes | — | Structure block: longer machined bed and ram, more stress-relief and handling, multi-point crowning |
| Control and tooling content | Comparable | Comparable — this is the block that does not move |
| One 4 m part | Two setups, or a tandem arrangement | One setup |
What is welded in, and what a purchase order can still change
| Item | Fixed at the factory? | Buying rule |
|---|---|---|
| Bending length | Yes — the frame is one weldment | Buy the length your longest repeat job needs |
| Throat depth | Yes — it is frame geometry | Buy it for the deepest return flange you actually form |
| Opening height and stroke | Yes — structural | Buy it for the tallest tooling stack you will ever mount |
| Tonnage and frame | Yes | Size it from your heaviest real job, not the brochure table |
| Synchronization architecture | Effectively yes | Decide on a five-year horizon, not this quarter |
| Controller and axis count | No — controller swap, added axes, licence | Specify what pays back now, upgrade when the work arrives |
| Crowning | No — retrofit kits exist for most frames | Add it if you bend parts beyond about 2 m to tight angles |
| Tooling and clamping | No — a purchase order | Buy for the parts you run weekly, not the exception |
| Guarding | No — retrofit is normal | Treat as a compliance item, not an option |
The rule that follows: length, throat depth, opening height and tonnage are decided by steel and cannot be revisited; control, crowning, clamping, tooling and guarding are decided by a purchase order and can. Never compress a dimension you will need for ten years in order to add an option you can add next year.
What the Price Does Not Cover
A quotation is bounded by its terms, and buyers routinely compare numbers with different boundaries.
Validity is arithmetic, not pressure selling. A press brake is quoted today and built later, out of steel bought later. In an interview published in Modern Metals in September 2026, Marc Lerman, chief commercial officer of the steel processor Steel Warehouse, described a market where a USD 50 to 60 per ton change can happen inside one week, with mill lead times of about eight weeks for hot-rolled and nine for cold-rolled and coated, and spot availability close to nonexistent (Modern Metals, 21 September 2026). Frame steel, tooling steel, the reservoir, the guarding, the container and the currency all have to be committed before the machine ships. So put four things side by side rather than one: price, validity, delivery term in Incoterms 2020 language, and lead time with its starting point — a lead time counted from order confirmation and one from deposit receipt are different promises.

Landed cost is not machine cost. The following sit outside a machine price and you pay for them regardless:
- Freight from the factory to your door, and marine insurance
- Port charges, terminal handling and documentation at both ends
- Customs duty, and any certification or inspection your country requires on import
- Inland transport from the port, plus a low-bed trailer if the machine is long
- Unloading and positioning: a crane or forklift rated for the machine’s weight, often rented by the day
- Foundation work, levelling pads or anchor bolts, and floor preparation
- Electrical supply to the machine, including a transformer if site voltage does not match
- First fill of hydraulic oil, where it is not shipped with the machine
- Commissioning, and the travel and accommodation of the engineer who does it
- Training, including the days your people are not producing
- A first spare-parts order, and the wait before a second one arrives
Two consequences. The comparison flips more often than buyers expect: an ex-works quote plus freight, duty, unloading and commissioning regularly lands above a delivered quote that looked higher on paper. And the only honest comparison is one where every supplier quotes the same delivery term and the same scope of supply, stated explicitly rather than as “delivered”.
Support is a line item even when it is not priced. A press brake is a ten-year asset, and the difference between a supplier who stocks spare parts and consumables for your model and one who orders them per incident is measured in weeks of lost production, not in the purchase price.
What a Low Quote Leaves Out
A low price is a statement about scope, not about the market. Something has been removed, and the useful question is what, and when you will meet it again.
| Where the saving is taken | Why it is invisible at the buying stage | How it returns within the first few years |
|---|---|---|
| No stress relief, or natural ageing instead of a controlled process | The frame looks identical once painted; it is not in the specification sheet | Angle drift on long parts as the frame relaxes — shimming, rework, arguments with your customer |
| One linear guide and a cheaper screw on the backgauge | Axis count and travel look the same on paper | Repeatability per shift; every batch needs a trial bend to re-establish position |
| Different brands of pump, valve and electrical components | Nameplates are inside the cabinet | Waiting weeks for a part, on a machine that earns money by the hour |
| A basic numeric control instead of a graphic one | Both are touchscreens, both are called CNC | Programming and sequence time per job, every shift, for the life of the machine |
| Tooling on general-purpose steel, unground, mixed segment lengths | The machine still bends metal on the first day | Die marks, steps between segments, longer changeovers, shorter tool life |
| Minimum guarding | The price difference is small and the line item is easy to skip | Reduced operating speed, or a compliance problem in the market you sell into |
| Support limited to the factory gate | It is the least visible line in the quote | The first failure becomes a project instead of a repair |
Seen on paper, the same gap usually looks like this:
| Supplier A | Supplier B | |
|---|---|---|
| Machine | 170 t / 3,200 mm | 170 t / 3,200 mm |
| Control | Basic numeric | Graphic, with offline programming |
| Backgauge | 2-axis | 4-axis |
| Crowning | Manual | CNC-controlled |
| Clamping | Manual | Hydraulic |
| Tooling | Starter set | Full package for your part range |
| Quoted price | Lower | Higher |
Neither column is automatically better. For one product, generous tolerances and a maintenance team on site, Supplier A is the correct buy; for a shop running several profiles a week, B pays for itself in setup time. The mistake is taking the savings without knowing which ones you took.
You do not have to accept any of this on a supplier’s word, including ours. Ask for the frame stress-relief procedure as a document, not a sentence. Ask for photographs of the pump, valve and controller nameplates on the machine that will actually ship. Ask for the load rating per metre of the exact tool being supplied. Ask for a customer within driving distance running the same model on similar work, and call them. Ask what happens, in writing, if the machine fails to hold angle during acceptance. A supplier who answers those five in documents is a different proposition from one who answers them in adjectives.
Turning a Price Difference Into Billable Hours

Take the difference between the two quotes and divide it by your shop rate. The answer is no longer money; it is billable hours. That is the real size of the gap, and the only form in which a shop owner can judge it, because hours are what you sell.
Then ask what the more expensive machine returns per month in the same unit. A press brake returns hours in only three places: setup, first-piece scrap, and unplanned downtime. Setup is the largest and the most ignored — changeover and trial bending consume hours that never reach an invoice, and they scale with how many different parts you run per week rather than how many parts you make. First-piece scrap is the second: trial bends that end in the bin carry the material, the cutting and the labor already invested in that blank. Downtime is the third, and it hurts most because it takes the whole cell with it.
A worked shape, with placeholders to replace: suppose the gap is worth 900 billable hours at your rate. A machine that removes five hours of setup a week returns about 260 hours a year, so on setup alone the gap takes three and a half years to close. Add less trial-bend scrap and one avoided breakdown a year, and the same gap may close inside two years. If it does not, the expensive machine is the expensive choice, whatever it does on the demonstration floor.
Two cautions. Do not count theoretical cycle time as a return; count the hours the ram is actually forming parts, because a fast machine that sits idle while someone hunts for a tool has returned nothing. And do not count capacity you have no orders for.
| Cost per part | What belongs in it | How to estimate it without a finance department |
|---|---|---|
| Depreciation | Machine price over its working life | Price ÷ years, then ÷ parts per year |
| Cost of capital | Interest, or the return you give up | Your borrowing rate on the outstanding balance |
| Energy | Power drawn while working and while idle | Read the meter over a normal week, not the motor nameplate |
| Maintenance and spares | Consumables, filters, oil, wear parts, service visits | Last machine's annual spend, adjusted for age |
| Tooling amortisation | Tooling cost over the parts it makes | Tooling price ÷ expected tool life in strokes |
| Setup | Changeover hours × changeovers × shop rate | Time three real changeovers with a stopwatch |
| Scrap and rework | Material, cutting, labor already in the blank, plus rework hours | One month of your reject tally, costed properly |
Run this on the two quotes you are holding and the conversation changes. It stops being about which supplier is cheaper and becomes about which machine returns its difference inside your planning horizon.
What Is Worth Paying Extra For, in Order
Pay for what the machine does every day, not for what it does once.
Crowning, if you bend long parts. Deflection is not a defect; it is what happens to any machine under load. Beyond about 2 m with a tight angle, a crowning system is the difference between bending a part and then fixing it.
Backgauge axes, if you run many flanges. The backgauge configuration is where setup time leaks. X alone handles simple flanges; R and Z1/Z2 decide how fast a multi-flange panel is set up, and repeated manual re-setting is the most common hidden cost in high-mix work.
A graphic control and offline programming, if the part mix changes. The CNC control system earns its money in programming and sequence time, and in not needing a veteran at the machine to work out the order of bends. Training belongs to this line: one bending-equipment maker’s product manager notes that customers with no previous experience can run their systems independently after a five-day course, which tells you that training days are a cost to specify rather than a line to leave open (Shop Metalworking Technology, 23 June 2025).
Clamping and tooling storage, if changeovers are frequent. A clamping system and organized tooling storage do not make better parts. They make more parts, because they remove minutes from every changeover, several times a day.
Guarding, always. This is not an option to be traded against price. Decide the arrangement and the standard the machine is built to before you discuss cost, and treat safety arrangements as a specification item with a document behind it.
What is usually not worth the money: tonnage you cannot use, stroke and opening height far beyond your tallest tooling stack, and any feature whose value depends on work you do not have. A machine bought for the exception spends most of its life being the wrong size for the daily job.
How to Request a Quotation You Can Compare
Send one document to three suppliers. It should carry:
- Material — mild steel, stainless, aluminium or other
- Thickness — minimum, typical and maximum
- Longest regular bending length
- Part drawings or CAD files showing the bending profiles
- Production volume — parts per shift, day or month
- Angle and dimensional tolerance
- Preferred drive and control, if already decided
- Backgauge axes, clamping and automation you expect
- Delivery term, in Incoterms 2020 language
- Destination country and site
Ask each supplier to mark every line included, excluded or optional, with prices separate, and to state the validity period. Then request a trial bend on your own part as the acceptance test. Quotes built on the same document are comparable; quotes built on three different sales conversations are not.
Where CAMT's Range Sits
We build these machines, so read this as a map rather than a recommendation.
Our range separates along the lines the cost blocks do. The Economical series uses mechanical torsion-bar synchronization with an NC control in a 2 to 3-axis configuration — the band where structure and power dominate the price and the control package is deliberately kept simple. The Smart series moves to synchronized 3 to 4-axis control. The Genius series adds 4 to 6 axes with motorized backgauge positioning and multi-point crowning, which is where the control block starts to carry real weight in the invoice. Above that sit the Apex hybrid 6 to 8-axis, the EPB servo-electric, tandem and heavy-duty configurations, each of which changes the power and drive block as much as the control block. Tonnage and dimension tables are on the press brake overview, and the tooling interface question is covered in the press brake tooling guide.
How we quote, so you can compare us against anyone else: we price against your part list, not against a model number. The quote comes back with the axis configuration, the controller model and its axis letters, the tooling scope, the guarding arrangement, the delivery term and the validity period stated separately, and acceptance is a trial bend on your part measured along the full length. If a supplier will not price against your parts, the number they give you is not about your work. Send the list through the contact page and ask for those headings back in writing.
This is one screen of a longer buying process. The press brake buying guide runs from part mix and tonnage through to the purchase order.
Frequently Asked Questions
How much does a press brake cost?
There is no single figure, and any number quoted without a specification is meaningless. What you can rely on is the relative shape: within one tonnage, configuration choices move the price more than the choice of builder. Price your own part list with three suppliers, and convert the differences into billable hours before deciding.
Does a higher tonnage always mean a higher price?
Within one architecture, yes, and faster than proportionally, because frame sections, cylinders, pump and valve sizes step up together. Across architectures, no: a well-specified smaller machine can cost more than a larger basic one. Force is only one of five blocks.
Does the quotation include tooling?
Not necessarily. Some quotes include a standard punch and die set; others list tooling separately. Confirm quantity, length, standard, hardness, maximum load per metre and whether special profiles are extra. Tooling you must buy later is still part of the purchase price — it is simply paid in a second invoice.
Is a servo-electric press brake more expensive to own?
It has a different cost structure rather than a uniformly higher one: no hydraulic oil to change, different maintenance items, and power drawn in relation to work done rather than at idle. Whether that pays back depends on how many hours the machine runs and how much of that time it is forming parts. The drive decision itself is covered in our servo-electric vs hydraulic vs hybrid comparison.
Should I buy a used press brake instead?
Only after you price the missing pieces: tooling, control support, spare-parts availability and the hours needed to bring it to your tolerance. A used machine also raises a standards question no brochure answers. ISO 6909:2026 states that it is not applicable to press brakes manufactured before its publication date, and what that means for a specific machine in your market is a question for your safety adviser and your local regulator, not for a supplier.





