metriqone real federation, measured end to end
Upload a Navisworks NWD, choose what belongs in the bid, and download a line-based MTO in your client’s own schema.
You can open an account and subscribe here. Reading this page needs no account.
Why take-offs go wrong
A federation stacks source models on top of each other — superseded revisions, duplicated geometry, the same run delivered by two disciplines. Walked by hand, every copy reads as new material.
Measuring a large model manually takes days, and those days run inside the tender window, not after it.
Procurement cannot buy from one number. They need the quantity split by code, nominal size and material — which is also how your client’s schema is laid out.
Three steps
Drag the NWD in — up to 300 MB, straight from your browser to processing.
~40 min end to end for a 210 MB federation · you can close the tab
Every source model in the federation is listed with its own quantities. Tick what belongs in the bid; the screen, the Excel and the PDF all follow.
under 2 minutes
Excel workbook and evidence report, in your client’s schema.
immediate
End to end, that is about 40 minutes on the largest federation we have run.
Deliverables
Summary, Heading Breakdown, MTO, Fitting Breakdown, Steel, Valves, Material Summary, Material Detail, Unknown Items, Method. Sheets appear when the model carries the data for them. Editable, in your client’s column order.
A printable pack behind the totals: diameter breakdown, fitting breakdown by code and size, what was excluded and why, and the method used. Hand it to your client with the bid.
Include or exclude any source model in the federation with one tick. One decision, applied to the screen, the workbook and the report at the same time.
Quantities split by material where the model publishes it — carbon, stainless, and the rest — so procurement can price directly from the sheet.
Click any row in the take-off and see the exact objects behind it, isolated in the model. Every quantity is traceable to geometry. Available on cloud-processed models.
Ask in plain English — “how much 4-inch pipe on the CIP lines?” The answer is executed against the take-off rows, not generated.
One real federation · anonymised
Five things happened between the upload and the number. Here is all of them.
A single federation lands on the desk. One file, every discipline, 210 MB. Somewhere inside it is every metre of pipe you are about to price.
1,300,000objects in one federation
The same runs are in there twice — a superseded revision here, geometry delivered by two disciplines there. The copies sit exactly on top of each other, so the model looks perfectly normal.
The eye cannot see it. The bid pays for it.
Every object is fingerprinted before it is measured. Byte-identical copies across source models collapse to one. Nothing is estimated, nothing is thrown away — the same pipe is simply bought once.
82,200duplicate objects counted once
7,104mof phantom pipe that never reached your bid
The full model is not the tender. Tick the source models that belong in the bid and the total moves with them — on screen, in the workbook and in the report at the same moment.
20,37610,692mfull model, then tendered scope
Open any line and the objects behind it light up in the model. Row, source model, geometry — the chain never breaks, so your client can check the same line you did.
Every row → its source model → the objects behind it.
Numbers come from a deterministic engine. AI never writes a quantity.
Client, project and file identifiers are never published. Anonymised figures only.
The system brain
Metriq keeps its take-off rules where you can read them — a living rulebook, written in plain sentences, in English and Turkish.
Each rule card records what happened and what it cost — a tee family that lied about its size, a generic coupling that was never a purchased item, 77 crosses a name map silently dropped. The rule states the fix; the incident stays attached as the reason.
Add a rule the way you would say it to a colleague — "drain lines below 2 inches are out of scope" — in English or Turkish. Toggle any rule off, or remove it; nothing needs a developer.
The engine enforces the measured rules and the AI auditor reads the whole rulebook on every new model — so a lesson learned on one client’s file guards all the files after it. No profiles to pick, nothing to configure per upload.
Pricing
A subscription at £200 a month for teams that measure every month, or a single model measured once for £100 — both excluding VAT, both paid by card in a few minutes. Talk to us first if you would rather.
Subscription
★ EVERYTHING INCLUDED£200/ month · excl. VAT
Unlimited users, 15 models a month, and the take-off comes out in your client’s schema — not ours.
Card at the next step, cancel any time. Prefer a written quote first?
Single file
£100one-off · excl. VAT
One NWD federation, measured once. No renewal, no monthly allowance to think about — you pay for the file in front of you.
Card at the next step. Nothing is charged again unless you buy another file.
Subscription: monthly, cancel any time from the billing portal — it ends at the close of the period you have already paid for, never mid-month.
Your first subscription month includes mirroring your MTO template — two to four days of work on our side, at no extra charge. We start it once you subscribe.
Need more volume? 25 models and 5 large federations a month — just ask.
Single file: refunded in full within 14 days if the model has not been processed. Once it has been processed, the purchase is final. Re-processing the same file after an engine update is included within fair use; template mirroring comes with the subscription.
Prices exclude VAT. Paddle adds any VAT due at your country’s rate and shows the total before you pay — a UK customer pays £240 a month, or £120 for a single file.
Charged to your card by Paddle, our merchant of record — monthly for the subscription, once for a single file. Paddle emails a VAT invoice after every payment and keeps them all in your billing portal — add your company name and VAT number at checkout and they appear on it.
Before you ask
Four things worth knowing before the first model.
We read what the model publishes. A federation of plain meshes or dumb solids carries no sizes or classes, so it produces no take-off rows — we tell you that instead of estimating.
That is the current ceiling for a single NWD. Larger federations are split by discipline and measured as separate runs.
Your client’s conventions are not ours — whether 45° bends fold into 90s, whether collars count 1:1, whether companion flanges are in scope. On a subscription we mirror your template in the first month, and it then applies to every model after it. A single-file purchase is measured with the rulebook and schema as they stand.
Calibrating and scoring on the same file is not a measurement. Instead you get the evidence: every row, its source model, and the geometry behind it. Check any line you like.
Send one federation. You get the Excel, the evidence report, and a walkthrough of every difference against your own numbers.
We reply within one working day.
