A Guide to LOD for Revit Models

If a Revit model arrives looking polished but cannot support the decisions your team needs to make, the problem is rarely Revit itself. More often, it is a mismatch in expectations around scope, geometry and usable detail. That is exactly why a clear guide to LOD for Revit models matters - not as a box-ticking exercise, but as a way to reduce risk before design work starts.

For architects, technologists and consultants working from existing buildings, LOD is not just about how much detail appears on screen. It affects coordination, measured accuracy, model usability, programme, and fee. Asking for the wrong LOD can leave you paying for model content you do not need, or worse, relying on information that is too light for the stage you are in.

What LOD means in Revit models

LOD usually refers to Level of Development, although some teams still use it loosely to describe Level of Detail. In practice, the distinction matters. Detail is visual. Development is about how reliable and decision-ready the modelled information is.

In existing-condition BIM, that difference is especially important. A wall drawn with clean geometry may look convincing at first glance, but if it has not been modelled with the level of dimensional confidence needed for design coordination, then its value is limited. A good Revit model is not simply tidy. It is dependable for the purpose it was commissioned for.

That is why LOD should always be discussed alongside intended use. Early feasibility, planning, coordination, listed-building intervention and fabrication support all call for different levels of model definition. There is no single best LOD. There is only the right LOD for the job.

A practical guide to LOD for Revit models

The most common LOD ranges in measured building and scan-to-BIM workflows are LOD100 through to LOD400. Not every project needs the full range, and many existing-building models sit comfortably between broad categories depending on brief, budget and building complexity.

LOD100 and LOD200

LOD100 is conceptual. It is useful when massing, overall form or broad spatial understanding is enough. For existing buildings, this level may suit very early option studies where teams need approximate volume, floor areas or simple context rather than room-by-room coordination.

LOD200 moves into generalised systems and elements. Walls, floors, roofs, doors and windows are represented more clearly, but still without the depth of information needed for detailed construction decisions. This can work well for planning-stage models, basic layout development, or projects where speed matters more than fine component definition.

The trade-off is straightforward. LOD100 and LOD200 are faster and more economical to produce, but they should not be mistaken for detailed record models. If your team needs to set out interventions against irregular existing fabric, these levels may not carry enough confidence.

LOD300

LOD300 is often the practical baseline for serious design work in existing buildings. At this level, the model geometry is sufficiently developed for coordination, measured reference and general design decision-making. Elements are located, sized and represented in a way that allows the model to function as a dependable design tool rather than a broad visual guide.

For many refurbishment, extension and retrofit schemes, LOD300 is the point where value and usability align. Architects can develop proposals with greater confidence, consultants can coordinate against a consistent base, and teams spend less time second-guessing what the surveyed model is actually telling them.

This is also where specification matters most. One client’s LOD300 may mean accurate wall positions and floor levels only. Another may expect ceiling features, structural framing, stair geometry, window recesses and roof build-up to be included. The headline LOD label helps, but the deliverables schedule is what prevents ambiguity.

LOD350 and LOD400

LOD350 starts to include more defined interfaces between building elements. It supports stronger coordination where the relationship between components matters, particularly in technically complex spaces. In existing structures with awkward junctions or non-standard geometry, this can be very useful.

LOD400 moves into fabrication-level intent. For many measured survey projects, this is only required in selected areas rather than across an entire building. If a joinery package, staircase, feature façade or specialist heritage element needs highly developed modelling, then a targeted LOD400 scope may make sense.

Requesting LOD400 for an entire existing building, however, is often unnecessary. It increases modelling time and cost, and in many projects the extra content does not improve the actual decision-making process. A more efficient route is often a mixed-LOD model, with core building fabric at one level and high-priority zones taken further.

How to choose the right LOD

The most reliable way to choose LOD is to start with use, not appearance. Ask what the model needs to support in the next project phase. Is it for concept design, planning, consultant coordination, heritage consent, technical design, or a contractor package? The answer should shape the brief.

It also helps to consider who will use the model internally. A design team testing layouts may need consistent geometry, levels and opening positions, but not every decorative feature. A conservation architect working on listed fabric may need more nuanced modelling of irregular walls, vaulted ceilings or historic joinery because those elements directly affect intervention decisions.

Budget and programme should be part of the conversation as well. A higher LOD is not better by default. It is only better when the additional information will be used. Precision-first documentation means matching output to purpose, not inflating scope.

Why existing buildings complicate LOD

New-build BIM discussions often assume repeatable components and predictable geometry. Existing buildings are different. Floors slope. Walls lean. Openings have been altered over time. Roof structures rarely behave as neatly as standard families suggest.

That is why a guide to LOD for Revit models needs to address measured reality, not just modelling standards. In heritage properties, converted buildings and architecturally sensitive spaces, the challenge is not merely how much to model. It is how to represent irregular conditions in a way that remains usable inside Revit.

There is always a balance to strike between geometric faithfulness and model performance. If every surface variation is over-modelled, the file can become heavy and inefficient. If everything is overly simplified, important conditions disappear. The right approach depends on what the project team needs to interrogate.

A dependable scan-to-BIM workflow usually resolves this by agreeing tolerances, content scope and modelling conventions at the outset. That keeps the model aligned with the point cloud and avoids unnecessary interpretation later.

What to specify beyond the LOD number

A single LOD label is not enough to define a useful model brief. To avoid rework, specify the building areas included, the elements to be modelled, the expected geometric accuracy, and any exclusions. State whether the model should include ceilings, primary structure, roof timbers, floor build-ups, window boards, stairs, MEP features or heritage detail.

You should also define the intended outputs. If the Revit model will support plans, sections, elevations and reflected ceiling information, that affects how the modeller approaches categorisation and content. If the model is mainly for spatial coordination, that may point to a leaner setup.

For complex or listed buildings, it is sensible to identify priority areas early. Not every room needs the same level of attention. Concentrating higher development where design risk is highest often gives better value than spreading detail evenly across the whole project.

Common mistakes when requesting LOD

The first mistake is asking for the highest LOD available without a clear reason. That often creates extra cost without improving design certainty.

The second is treating LOD as a guarantee of universal accuracy. A model can be LOD300 and still exclude specific elements if they were outside the agreed scope. That is not a failure if the brief was clear, but it becomes a problem if assumptions fill the gaps.

The third is overlooking communication during mobilisation. Existing-condition modelling benefits from early dialogue, especially where geometry is irregular or the building has multiple phases of alteration. A brief conversation at the start can prevent significant revision later.

LOD as a risk-management tool

For design professionals, the real value of LOD is operational. It helps align survey capture, modelling effort and downstream use. When that alignment is right, teams move faster because they trust the base information. When it is wrong, coordination slows and confidence drops.

That is why experienced documentation partners do not simply ask, “What LOD do you want?” They ask what decisions the model needs to support, how the building behaves, and where project risk sits. For firms working across England and Scotland on refurbishment, heritage and complex-geometry projects, that conversation is often the difference between a model that looks acceptable and one that genuinely supports design.

If you are specifying a Revit model for an existing building, aim for clarity over headline labels. A model that is accurately scoped, sensibly developed and easy to work with will always outperform one that sounds more detailed on paper but misses the real needs of the project.