Planlift logoAll articles
3D Modelling5 min read

Interactive 3D Models in Planlift: From Drawings to Orbitable Sites and Real World Context

How our new 3D model engine turns floor plans and elevations into interactive models, unlocks camera angles from street level to aerial, and places proposals directly into Google 3D maps.

Single image visualisations have changed how fast teams can test ideas. You drop in an elevation or a site layout, and within a couple of minutes you have a visual that brings a scheme to life. But a flat image is locked to one camera angle. If you need a view looking down the access road, a view from the neighbouring garden, and an aerial perspective showing the roofscape, you have to generate separate images and hope they all imagine the same building.

With our latest update, you can start those views from the same model. We have launched our biggest feature release yet: interactive 3D Models in Planlift.

Now you can upload your site drawings and turn them into an orbitable 3D model. Once your model is ready, you can move the camera, hit Render this view to create visuals from different perspectives, and place the proposal into real surroundings in Google 3D.

Here is how the new workflow works and what it means for your projects.

1. Build an interactive 3D model directly from drawings

Instead of building geometry by hand, you upload the drawings you already have on file: PDFs, images, or supported CAD drawings. Floor plans, elevations, and site layouts help the AI understand the building and how it fits together.

Before generating, you choose the level of detail that fits the current stage of your project and review the credit price.

Basic massing creates simple site layouts and building volumes with heights and roof forms, without doors or windows. It is useful for exploring larger schemes at an early stage. For a single building, the choices are Balanced and Highly detailed.

Balanced builds out individual house types with doors, windows, porches, chimneys, and coloured materials, placed on a site layout. Highly detailed adds stricter checks against the supplied elevations for more developed visuals.

Behind the scenes, a dedicated managed AI agent reads your plans and builds the model in an environment with 3D modelling software. You do not need to keep your browser tab open while it works. The build runs in the background, with an email and an in app notification when it is ready.

2. Render any perspective you want

Once your model is ready, you can choose the camera views that matter to your project.

Open the interactive viewer to rotate, pan, and zoom. You can move towards eye level along the street frontage, tilt into a front three quarter perspective, or pull back for an aerial view. When you find the viewpoint you need, click Render this view.

Planlift takes that camera view into its image creator so you can turn the model into an architectural visual. If you need several views of a scheme, orbit to each angle and render it. Each visual starts from the same model rather than asking AI to infer the building again from a flat drawing. Image generation still needs your review, as details can vary.

3. View your scheme on site in 3D

Showing a proposed development in isolation only tells half the story. One of the first questions planning officers, committee members, and clients ask is how the proposal relates to its neighbours.

We have connected Planlift to Google 3D photorealistic mapping. You can position, rotate, and scale your model in its geographical setting, then explore how the scheme appears against nearby buildings, trees, and terrain.

Because the model exists in three dimensions, you are not just pasting a flat drawing over satellite imagery. You can move around the proposal in context. This is an indicative view, not a measured assessment, and existing structures in the map may overlap your model.

Google mapping is for viewing only. Planlift does not render, capture, or export the Google imagery, and it must not be saved through screenshots. Render this view uses your model in the separate model viewer, not the Google map.

4. Free uploads for your existing 3D files

Many practices already build preliminary models in SketchUp, Revit, Rhino, or other design tools. You can bring that work into Planlift by exporting to a supported model format.

Planlift accepts existing 3D files including GLB, glTF, FBX, OBJ, STL, DAE, 3DS, PLY, 3MF, and IFC. Uploading your own model is free. You can then explore it in the viewer, position it on Google 3D, or create architectural renders. Image renders use credits as usual.

5. Download and export your model

Your models are not locked inside Planlift. You can download GLB and, where returned by the AI builder, the Blender project file. The viewer also offers OBJ with materials, STL, and glTF exports.

These mesh formats let you take the geometry into other compatible software. They are not a substitute for a fully authored BIM model, and support for materials and model structure varies between applications.

Tips for best results during Beta

The 3D model builder is currently in beta. A few practical steps help it interpret your drawings more reliably.

First, upload clear, legible drawings with labelled plots and identifiable house types. Include corresponding floor plans and elevations so the AI can match each building to the right drawings. Dense hatching, unclear labels, or missing views can make that interpretation harder.

Second, start with schemes of 60 units or fewer. This is the current scheme size limit, and smaller, clearer drawing packs are a useful starting point.

Third, remember that this is an indicative tool for visualisation and early stage illustration. The AI can make assumptions when piecing together a building from drawings. Always check the model and its renders against your plans. Do not rely on them for dimensions, construction, or verified technical accuracy.

Log into Planlift, open the 3D Model tab on your home screen, and try it with your own drawings or an existing model.

Made with Planlift. AI generated indicative models and visuals for concept and communication purposes only.