Quantity takeoff is one of the most repetitive, time-consuming, and critical tasks in the construction process.
Enough to participate in a race dcontract to understand how much time it takes: each partecipazione alle gare dcontract requires theextractio
Anyone who has prepared a bid knows how much work it involves: every tender requires extracting quantities from drawings and BIM models, verifying the results, and transferring them into cost estimates—often under tight deadlines.
Yet, when a project is available as an IFC model, most of that information already exists. It is structured, organized, and machine-readable.
In this article, we’ll explore why IFC quantity takeoff is still a bottleneck, the different approaches available to automate it, and how deterministic artificial intelligence is changing the way estimators and BIM professionals work.
What is IFC quantity takeoff?
IFC quantity takeoff is the process of extracting project quantities—areas, volumes, lengths, counts, and weights—directly from an IFC (Industry Foundation Classes) model, the open standard developed by buildingSMART for exchanging BIM data.
Unlike traditional quantity takeoff based on 2D drawings, model-based quantity takeoff does not measure drawings. Instead, it queries the model’s database, where every object—walls, doors, columns, rooms, ducts—contains geometry together with semantic information and properties.
In theory, this means quantities should be available with only a few clicks.
In practice, things are rarely that simple.
Why IFC quantity takeoff is still a bottleneck
There are four main reasons why extracting quantities from IFC models remains a slow and specialized task.
Every request is different
A request such as “Net wall volume by floor” requires a completely different calculation than “Total skirting board perimeter for all rooms.”
Each query involves different filters, relationships, measurement rules, and calculations. Every new request often means creating new schedules, configuring filters, or writing scripts.
These types of queries go far beyond what a simple IFC viewer can provide and usually require BIM authoring software or advanced model-checking tools—skills that many estimating departments simply don’t have.
Every model is different
No two IFC models are exactly alike.
Models created by different design firms or disciplines often use different classifications, naming conventions, property sets, and modeling practices.
Before quantities can even be extracted, estimators frequently spend valuable time understanding how the project has been modeled.
Results must be verifiable
Extracting a number is only part of the job.
Professionals need to know exactly which objects were included in the calculation, according to which rules, and they must be able to reproduce the same result consistently.
Without traceability and repeatability, quantity takeoff cannot be trusted for procurement or contractual documentation.
The data already exists—but remains difficult to use
This is the paradox of BIM.
The information already exists inside the IFC model, yet transforming it into reliable quantity takeoff often requires days of work by experienced technical professionals.
For bid departments evaluating multiple tenders every month, this directly limits productivity and the number of opportunities they can pursue.
Why generic AI isn’t enough
Given today’s rapid adoption of AI, it’s tempting to ask a general-purpose chatbot to calculate quantities from an IFC model.
Unfortunately, that approach is fundamentally flawed.
Large Language Models are probabilistic systems. They generate the most likely answer—not necessarily the correct one.
Applied to quantity takeoff, this means they may produce numbers that appear plausible but cannot be verified.
There is no guarantee regarding:
- which objects were considered,
- how the calculation was performed,
- whether the result is repeatable,
- or whether mistakes can be identified and corrected.
For documents that become part of construction bids, cost estimates, or project validation, plausible is simply not an acceptable standard.
Deterministic AI: AI that executes instead of guessing
A different architecture is needed.
This is the approach behind EDBIM.
Rather than calculating quantities directly, AI interprets the user’s request written in natural language and activates specialized software tools (MCP Tools) that perform the actual calculations directly on the IFC model.
Because the calculation is performed by deterministic software—not by the language model—the outcome is always identical for the same model and the same question.
The result is:
- deterministic,
- traceable,
- repeatable,
- and fully verifiable.
This is the principle behind AskIFC, EDBIM’s AI assistant for IFC models.
A typical workflow looks like this:
- Upload the IFC model—no BIM authoring software required.
-
Ask a question in natural language, such as:
- “Give me the net volume of all walls grouped by floor.”
- “List all doors narrower than 800 mm.”
- AskIFC activates the appropriate calculation tool, queries the model, and returns an exportable table containing both the results and the list of objects involved.
- Save the query and reuse it whenever needed.
The same library of queries can then be executed automatically across every model in a project—or every new design revision.
Instead of rebuilding calculations every time, organizations gradually create a reusable knowledge base containing recurring questions such as:
- floor areas by occupancy,
- door and window quantities,
- MEP network lengths,
- parameter completeness checks,
- room schedules,
- quantity reports,
- and much more.
What once required several working days becomes a standardized workflow that runs in minutes.
Benefits for estimating and bid teams
For organizations preparing bids, automating IFC quantity takeoff delivers measurable improvements in three key areas.
Faster delivery
Standard quantity reports can be produced in minutes rather than days.
Technical staff can focus on activities that actually create value, such as pricing strategies, risk analysis, and proposal optimization.
Greater accessibility
Natural language queries eliminate the need for advanced BIM expertise.
Even professionals who are not BIM specialists can extract information from IFC models without relying on technical departments for every calculation.
Better reliability
Every quantity can be traced back to the model objects that generated it.
Calculations are fully repeatable.
Whenever the IFC model is updated, the same saved queries can simply be executed again, immediately producing updated quantities and highlighting the differences between revisions.
The same approach also benefits public owners and contracting authorities, enabling them to verify submitted IFC models without purchasing expensive software licenses or outsourcing every validation task.
Beyond quantity takeoff: model validation
Reliable quantity takeoff depends on reliable models.
For this reason, quantity extraction naturally goes hand in hand with model validation.
Before relying on model quantities, organizations should verify that IFC models are complete, consistent, and compliant with project information requirements.
This includes checking:
- required properties,
- parameter completeness,
- geometric clashes,
- model consistency,
- version differences,
- and IDS compliance.
This is precisely the role of CheckIFC, EDBIM’s solution for clash detection, model comparison, and IFC validation based on custom rules and IDS files.
We explore these capabilities in detail in our dedicated article on IFC model validation, and we’ll demonstrate them live during our autumn webinars and at SAIE.
Frequently asked questions
Do I need BIM authoring software to perform quantity takeoff from an IFC model?
With traditional workflows, yes. BIM authoring or dedicated model-checking software is typically required.
With an AI-first platform such as EDBIM, you simply upload the IFC file and ask questions in natural language—no authoring software or specialized BIM licenses are needed.
Are AI-generated quantities reliable enough for official cost estimates?
It depends on the underlying architecture.
Pure generative AI cannot guarantee repeatability.
A deterministic system like EDBIM performs calculations through specialized software tools, ensuring that identical inputs always produce identical results while maintaining full traceability back to the original model objects.
Can I calculate quantities across multiple IFC models?
Yes.
Saved queries can be executed in batch across all models in a project or every newly delivered revision, turning quantity takeoff into a continuous process rather than a one-time activity.
Does automated quantity takeoff work with IFC models created by other companies?
Absolutely.
This is one of IFC’s greatest strengths.
Because IFC is an open standard, models can be queried regardless of the authoring software used to create them.
Naturally, the quality of the results depends on the quality of the information contained in the model—but that quality can also be verified automatically using dedicated validation tools such as CheckIFC.
