Revit 2027.2 adds a tech preview called Incremental Model Open. It’s a small checkbox, but it changes how Revit opens a model.

Why We Need Incremental Model Open

Today, Revit reads the entire model into memory when you open it. Every element, every view, every sheet. Even the parts you won’t touch today.

Before the first view appears, Revit reads the model data, rebuilds element relationships, processes links and imports (RVT, DWG, PDF, images, point clouds), and generates graphics. The bigger the model, the longer you wait. And the more RAM you need.

That is why many teams split models. Not only because the project needs it, but also because the hardware does. A large data center model ends up as several MEP models, just so each one opens in reasonable time.

Incremental Model Open takes a different approach. Revit processes only the data needed for your starting view. The rest is processed when you need it.

Autodesk claimed the initial tests show up to 1.5 to 2 times faster model open in some scenarios, and 60 to 80% less data processed at open. Your results may vary.

Autodesk calls it the first step of a bigger plan: working with subsets of a model and, later, streaming model data directly from the cloud.

Feature Availability

Incremental Model Open first released as part of the Revit 2027.2 Update, and available for all updates after that.

A few things to know before you try it:

• It’s a tech preview. Read the Autodesk Tech Preview Terms and Conditions when you turn it on. You may want to review it first before make it your team standard yet.
• It’s off by default. You need to enable it yourself.
• It’s an application setting. It applies to Revit, not to a model.
• It only speeds up opening.
• Cloud models still download the entire model. Revit still downloads and caches the whole cloud model before opening.

There are also known limitations, mostly around working with links. Read the details by opening “Learn more about Incremental Model Open” in the option or here: Incremental Model Open (Tech Preview).

Enabling Incremental Model Open

There are two ways to turn it on. You can use the Options dialog or modify Revit.ini file.

Option 1: From the Options dialog

Close all of your models. The setting can’t change while a model is open.

Open Options: Go to File > Options > General. Find the Incremental Model Open section.

Option 2: From Revit.ini

Open Revit.ini, under the [Misc] section, add this line:

IncrementalOpenDynamic=1

Start Revit: The feature is active the next session Revit starts. If you have Revit open, you need to close it then restart it.

Back up Revit.ini before you edit it.

Using it

Open your model as usual, all files opened after changing the option will use Incremental Model Open.

You should feel the effect when opening the first view. Opening another view may take a moment, because Revit processes data it skipped earlier.

Tip: Your starting view decides how fast the model opens. Set a lightweight starting view, like a simple drafting view or a project info sheet, in Manage > Starting View.

Summary

Incremental Model Open doesn’t make Revit faster. It moves work from the moment you open a model to the moment you actually need the data. For large models, that can mean a much shorter wait and less memory at startup.

It’s a tech preview, so it is resommendeed to try it on project copies first.

But some commands never made it to the ribbon. Others sit in places that’s hidden make nobody clicks. AutoCAD veterans still type them every day, because they solve problems that ribbon doesn’t solve faster.

Here are 10 worth stealing from them.

1. MULTIPLE

Some commands end after one action. Draw one circle, and you’re back at the prompt.

Type MULTIPLE, press Enter, then type the command name. AutoCAD repeats it until you press Esc. It works well for CIRCLE, POINT or INSERT when you need many in a row.

It repeats the command, not your input. Options must be entered each time.

And yes, you can press Enter or Space to activate last command too. However, MULTIPLE can be more convenient.

2. OOPS

You erased a group of objects. Then you drew a few lines and moved a door. Now you need the erased objects back.

UNDO would cancel all work you did. OOPS restores only the objects removed by the last ERASE. Everything you did after stays where it is.

It’s a neat trick to remove some objects that interfering when you draw, then call them back!

3. ‘CAL

Need to find a point 1/3 between two reference end points?

When AutoCAD asks for a point, type 'CAL, then type the formula (end+end)/3. The apostrophe runs CAL inside the active command, and the result goes straight into it.

CAL understands basic math if you need to calculate a distance. You can type 1200/3 when you need to calculate an offset value.

Not available in AutoCAD LT. LT users can use QuickCalc instead.

Find more about QuickCalc here.

4. FILTER

Quick Select is great, however real drawings often need more: all circles on layer S-COLS, or all text with a height of 2.5 that is not on layer A-ANNO.

FILTER opens the Object Selection Filters dialog box. You combine conditions with AND, OR and NOT, then save the filter to reuse it later.

Tips: type 'FILTER at any Select objects prompt. The filter command can run inside MOVE, ERASE or any other command, even with a crossing window.

Find more about FILTER here.

5. EXPORTLAYOUT

A client asks for the sheet “as a normal drawing”. No layouts, no viewports, just everything on one model space. While using layout is a best practice, the client requirement might be different.

Right-click the layout tab and choose Export Layout to Model, or type EXPORTLAYOUT. AutoCAD saves a new drawing with everything visible on that layout in model space.

Check the result before you send it. Linetype scale in xrefs and blocks may not be kept as the original drawing.

6. BURST

Explode a block with attributes, and you get the original attribute tags: PROJECT_NAME, DWG_NO, SHEET_NO. The actual values are gone.

BURST explodes the block but turns each attribute value into plain text, exactly as it appeared. It’s useful when you send drawings to people who don’t need your blocks.

You’ll find it on the Express Tools tab> Blocks panel> Explode Attributes.

Not available in AutoCAD LT.

7. ARRAYCLASSIC

The modern ARRAY command creates associative arrays and opens a contextual ribbon tab. Powerful, but sometimes you just want a simple quick array.

ARRAYCLASSIC opens the old Array dialog box. Set rows, columns and spacing, preview, click OK. The result is separate objects you can edit freely.

8. COPYBASE And PASTEORIG

Ctrl+C and Ctrl+V work between drawings, but you can’t choose the base point. Placing the copy precisely becomes a guesswork.

COPYBASE (Ctrl+Shift+C) asks for a base point first. Pick a reference point then the object. Now when you PASTECLIP or press Ctrl+V you can place the object using a base point.

If both drawings share the same global coordinates, use Paste to Original Coordinates (PASTEORIG) from the right-click Clipboard menu. The objects will be placed at the exact same coordinateas in the source drawing.

9. OVERKILL

Drawings from other parties often have lines drawn on top of lines. They break hatch boundaries, confusing when using snaps and offset. Too many overlapping lines also can inflate the file and make it prone to corrupt.

OVERKILL (Manage > Cleanup > Overkill) removes duplicate and overlapping lines, arcs and polylines. It can also combine collinear segments into one line.

10. UCS Object + PLAN

One wing of the building is rotated. Or the drawing follows the project orientation. Drawing on it with Ortho is painful.

First you need to change the UCS. Type UCS, then OB (Object – or you can pick points to define the UCS), and pick the object as reference. The UCS now follows that object.

To rotate the view type PLAN and choose Current UCS. The view rotates so the wing sits straight on your screen. Ortho, rectangles and polar tracking now follow the building orientation.

To go back, type UCS > World, then PLAN > World.

Old Commands, Real Problems

These commands are old but always relevant to everyday problem. Sometimes knowing a few commands by name can help you to be more productive.

Which old-school command do you still type? Share it in the comments.

1. Why a Healthy Model Matters

A model that opens may not be a healthy model.
Health shows in sync time, file size, warnings and stability. When it drops, it affects everyone, not only the BIM manager. Designers wait for syncs, coordination slips, and recovering a bogged-down model can take hours.
The causes are very common:

• CAD files imported instead of linked
• In-place families where a loadable family belongs
• Imported images and too many links
• Warnings nobody resolves
• Team members syncing with different minor versions, such as Revit 2027.1 and 2027.2

Most BIM managers know this list. The hard part is seeing it in time, without proper tool to analyze the problem.

2. What Is Model Analytics?

Autodesk Model Analytics (AMA) is a cloud-connected add-in for Revit cloud models. Each time someone syncs a workshared model, or saves a cloud model, it sends model data to a cloud service. The results appear as dashboards in Autodesk Forma Insight. So you don’t need to open Revit or run a separate tool to review them.

3. Who Can Use It?

Anyone who uses Revit 2024 to 2027 and works with cloud worksharing and/or cloud models can send the model health information to Forma.

To see the model analytics insight dashboard, you need to be given access to Insight. Project Members and Project Admins can view the project and model reports, read-only. Executive access is only needed for the hub-wide view. BIM coordinators and BIM managers will use it most, but the color-coded scores are simple enough for designers and project managers to quickly notice if there’s any issue.

4. How to Enable It

Model Analytics is off by default, and installing the add-in alone does nothing. Setup takes three steps:

1. Activate it for the hub: An Account Admin opens Hub Admin > Settings > Advanced and clicks Activate.
2. Install the add-in on every Revit computer: Download Autodesk Model Analytics for Revit from manage.autodesk.com (Products and Services > Revit > View Details > Extensions).
3. Sync the model: Data appears only after you Sync with Central or save the cloud model. All the models that you sync before you enable Model Analytics and with Revit without the extension will not have the data.

Then open the project in Autodesk Forma and go to Insight > Model Analytics.

5. What You Can Review

Reports come at three levels: one model, one project, or the whole hub.

Health Score: A score card built from the health indicators, from 0 to 10. Green is 7.8 and above, orange is 5.0 to 7.7, and red is below 5.0. A history chart shows whether the score is improving or getting worse.

Health Indicators: Each indicator is rated Acceptable, Caution or Needs Attention against default thresholds:

Yes, a single in-place family already means Caution! The thresholds can’t be edited, at least not yet.

Learn how to read the reports in detail in this Autodesk help article: Model Analytics.

The Bottom Line

Model Analytics doesn’t fix your model. It tells you where to look, and when. Someone still has to remove that CAD import. Someone still needs to keep the file efficient.

What changes is the habit. Model health moves from guessing to monitoring. Continuous model monitoring catches problems before the crash.

How do you track model health on your projects today?

What Changes on July 6, 2027

• Cloud access is limited to the current Revit release plus five previous versions.
• About 90 days after each new Revit release, the oldest version drops.
• Your data in the BIM 360 and Forma web apps, and local Revit work, are not affected.

You can see the schedule illustration below.

Each cut-off removes the oldest version from the window. It is recommended to upgrade to the latest version possible.

What to Do Before The Change

The most convenient way to upgrade cloud models is using Revit Cloud Model Upgrade tool.

1. Find active projects on Revit 2022 or older, that use cloud models or cloud worksharing. Finished projects can be upgraded later when necessary, even after the cut-off.
2. Agree on one target version with every firm on the project. Remember that upgrade can’t be undone. Autodesk recommends 2025 or newer.
3. Run a Test Upgrade, then book the downtime as nobody can save or publish while the project upgrades.
4. Check what the upgrade can’t process. Links through Desktop Connector and published copies in Design Collaboration or Bridge need upgrade from your computer.

If You Can’t Upgrade Your Revit Version

Download the cloud models and keep working locally in the same Revit version, without cloud collaboration (link files or local worksharing). Make the switch before July 6, 2027, while everyone can still sync.

The Bottom Line

The upgrade itself is a simple button click. The real work is coordination. For your next multi-year project, use the latest possible version if possible. One year before the latest release or latest release with at least has one update released is a common version to use.

References

If you need to find more details about this change, you can find them in the references below:

• Changes to cloud model access due to deprecation of older versions in Revit (Autodesk Support, updated 19 Oct 2026)
• Prepare for Revit Cloud Model Access Deprecation: Models Upgrade Process (Autodesk Support; covers workshared and non-workshared cloud models)
• FAQ: Understanding Deprecation of Revit Cloud Models Access for Older Revit Versions (PDF, October 2026)
• Cloud Models (Revit 2027 Help: Revit Cloud Worksharing vs Cloud Models for Revit)
• How to check the Revit version of a Cloud project

If a system follows predefined rules, that’s automation. It’s valuable, but it doesn’t involve intelligence.

AI comes into play when the system can interpret information, recognize patterns, make predictions, or generate responses beyond a fixed set of rules.

Automation can exist without AI. AI can exist without automation. Some solutions often combine both. Sadly, too much of today’s “AI” is simply automation with better marketing.

It won’t.

BIM is not a magic solution. It doesn’t eliminate design conflicts, reduce delays, improve coordination, or lower costs by itself. What BIM does provide is better information, visibility, and insight. But none of these solve the problem on their own. People solve problems. BIM helps people make better decisions.

Before asking, “How do we implement BIM?” it may be more valuable to ask:

• What problem are we trying to solve?
• What decisions need to be made?
• What information is missing today?
• How can BIM help us obtain that information?

When BIM initiatives fail to deliver value, it’s often because the focus is on creating models and fulfilling requirements rather than solving business problems.

Start with the problem.
Then identify the decisions that need to be made.
Then determine the information required to support those decisions.
BIM is not the destination. It’s a tool that helps you see more clearly so you can make better decisions.

Then things started to change

As BIM became more established, the industry developed a growing collection of frameworks, standards, classifications, and terminologies. We got BIM dimensions. We got Levels of Development (LOD). We got BIM maturity levels. We got execution plans, information delivery specifications, and countless acronyms. Instead of clarifying things, these sophisticated terminologies introduced a massive wave of confusion.

None of these concepts are inherently bad. Most of them were created with good intentions. The goal was often to improve consistency, clarify expectations, and help organizations scale BIM adoption.

But somewhere along the way, something interesting happened. The conversation slowly shifted. Instead of asking “how can we use this tool to solve a design issue?” and more about a existential crisis: “What exactly are we supposed to deliver to satisfy this specific acronym?”. Instead of focusing on outcomes, many discussions became focused on definitions. The focus quietly drifted from the practical value of the model to checking off boxes on a compliance spreadsheet. For many practitioners, BIM has become something they need to satisfy rather than something that helps them work.

Then came BIM mandates.

Governments, owners, and organizations around the world started requiring BIM on projects. Again, the intention was understandable. Better collaboration, better information management, and better project outcomes are difficult goals to argue against.

But mandates introduced a new challenge. Many teams were asked to implement BIM before they fully understood what BIM meant in practice. As a result, projects often found themselves in a strange situation. Teams were required to “do BIM,” yet there was no universal agreement on what success actually looked like. And because BIM became strongly associated with 3D modeling software, a common perception emerged: If it’s not 3D, it’s not BIM. That perception still exists today. Yet many project activities that support information management, collaboration, coordination, and decision-making do not necessarily require a detailed 3D model.

In some projects, the model creates tremendous value. In others, the model becomes an expensive obligation. There are projects where teams build highly detailed models because the contract requires it, even though very few stakeholders actually use the information. There are also projects where simple workflows, spreadsheets, drawings, and meetings solve problems more effectively than an increasingly sophisticated model.

This raises an uncomfortable question: Are we using BIM because it helps, or because we are expected to?

To be clear, I’m not arguing against BIM. Many of the industry’s most significant improvements have been made possible through BIM-related technologies and workflows. The question is whether the industry confuses the tools, standards, and deliverables with the original purpose.

What do you think?

• Has BIM become more useful over time, or more complicated?
• Do frameworks such as LOD, BIM dimensions, and maturity levels create clarity, or do they create confusion?
• Can a project be considered successful in BIM terms even if it relies very little on 3D modeling?
• And most importantly, when we talk about BIM today, are we still talking about helping people work better—or are we talking about something else entirely?

The discussion remains open.

Forma Board is a useful feature to communicate and explore ideas with your team. You can extract a thumbnail or a view from a file that you have in FDM and put that in a board and share it with the team. You can also embed an external image in your board.

Now it gets interesting that Forma Board now has the AI-powered rendering. You can render any view you extracted from Revit, IFC, or other 3D models in this board. The board can also create a photorealistic rendering from an image in this board.

The prompt has two options: precise and explore. The Precise option will keep geometry unchanged, and Explore will change the form. You can also choose standard or high quality for the two options.

Type the prompt to create a visualization from that view. The process is quite fast, even for the high quality option.

Now you can use Forma Board to explore the design ideas with your team and clients more effectively. Not only can you share views, images, and notes, but you can now also quickly explore a visualization of your design.

Interesting to see if you can do this directly in Revit soon, as now Revit is the Forma connected client.