RealityScan
Epic Games photogrammetry suite, formerly RealityCapture, free below USD 1M revenue and built for merging photos with laser and SLAM scans.
MeshLab
Open source system for processing and editing 3D triangular meshes, widely used for scan cleanup, repair and format conversion.
Overview
MeshLab is the open source workhorse of 3D scan processing. It began as a University of Pisa course project in 2005 and is developed by the Visual Computing Lab of ISTI-CNR, Italy's National Research Council, under researchers including Paolo Cignoni. It sits between raw 3D capture and production: it ingests dense point clouds and messy scanned meshes, then cleans, repairs, aligns, reconstructs, decimates and converts them into watertight models ready for printing, archiving or the web. It is GPL licensed and entirely free, its algorithms (like Quadric Edge Collapse decimation and Screened Poisson reconstruction) come straight from peer reviewed research, and institutions like the Smithsonian run it headlessly at industrial scale to digitise museum collections. The current release is MeshLab 2025.07, which added native Apple Silicon and Linux ARM64 builds plus 3MF support.
Key features
Quadric edge collapse decimation
The industry reference for polygon reduction: cuts multi million face meshes down to printable or web ready sizes while preserving boundaries, sharp edges and UV coordinates. Users regularly rate it above paid alternatives.
Surface reconstruction from point clouds
Turns unstructured scan points into watertight meshes with Screened Poisson reconstruction and Ball Pivoting, the standard step after photogrammetry or LiDAR capture.
Mesh cleaning and repair
Hundreds of filters remove duplicate and unreferenced vertices, fix non manifold edges, close holes and heal the defects that plague scanned data, so slicers and engines accept the result.
Scan alignment and texture projection
Aligns overlapping range maps with Iterative Closest Point plus global bundle adjustment, and registers photos onto geometry with a Mutual Information algorithm to build photoreal textured replicas.
A huge format translation matrix
Reads and writes PLY, STL, OBJ, glTF/GLB, Collada, X3D, U3D and more, plus E57 LiDAR point clouds and, since 2025.07, full 3MF for modern 3D printing. A common bridge for formats other tools reject.
PyMeshLab batch automation
The entire filter ecosystem is scriptable from Python: record a filter script in the GUI, then run it across thousands of files headlessly. This replaced the old meshlabserver command line tool.
Pros and cons
What users consistently praise
- Decimation quality that users rate above expensive commercial tools, with boundaries and UVs preserved
- Handles meshes of millions of polygons faster than many paid packages, without exotic hardware
- Converts between formats other software rejects, including E57 LiDAR data and legacy formats
- Completely free under the GPL, with research grade algorithms and Python automation via PyMeshLab
What users consistently criticise
- No undo at all: a wrong filter means duplicating layers beforehand or reloading the last saved file
- A cluttered, technical interface with deeply nested menus and parameter names aimed at geometry experts
- Extreme filters can hard crash the app, and with no autosave that means losing the session
- Official written documentation is thin, so learning leans heavily on community video tutorials
Recurring themes across third-party user reviews (G2, Capterra, TrustRadius).
Videos
Pricing
| Plan | Price | What you get |
|---|---|---|
| MeshLab | Free forever | GPL licensed, free for unlimited commercial, academic and private use. No tiers, no fees; the ISTI-CNR team accepts voluntary donations. |
| PyMeshLab | Free (pip install) | The Python library exposing MeshLab's filters for scripted, headless batch processing. Same license, same price: none. |
Best for
Heritage and research digitisation
Archaeologists, museums and labs clean, repair and archive artifact scans with it; the Smithsonian's digitisation pipeline runs MeshLab headlessly at industrial scale.
3D printing repair and decimation
Fix non manifold geometry, flip normals and crush a 10 million polygon sculpt into a size a slicer can chew, free.
Photogrammetry and scan pipelines
The standard post processing stop after capture: align range maps, reconstruct surfaces and project photo textures.
Engineers automating mesh work
PyMeshLab turns any workflow into a repeatable script, ideal for robotics, computer vision datasets and processing farms.
Who should look elsewhere
Beginners wanting simple edits
The dense, technical filter menus assume geometry knowledge. For casual mesh tweaks, browser tools or Blender are far friendlier.
Parametric CAD design
There is no solid modeling, no dimensional constraints and no NURBS. Engineers designing parts need SolidWorks, Fusion or FreeCAD instead.
Sculpting and animation
No sculpting brushes, no rigging, no timeline. MeshLab processes existing geometry; it does not create or animate it.
Frequently asked questions
Why is there no undo in MeshLab?
It is an architectural constraint, not an oversight. The underlying VCGlib framework trades state history for the memory efficiency that lets MeshLab chew through multi million polygon meshes. Duplicate the layer before experimental filters, or use Reload to fall back to the last saved file.
Is MeshLab really free for commercial use?
Yes. It is GPL licensed with no tiers, fees or restrictions on commercial output. Do not confuse it with MeshLib, a similarly named commercial SDK from a different company that charges licensing fees.
Can I batch process files without opening the GUI?
Yes, with PyMeshLab. Record a filter script in the GUI, then run it from Python across any number of files. The old meshlabserver command line tool was removed in 2020.12 and fully replaced by this library.
How do I shrink a huge STL for 3D printing without losing detail?
Use Quadric Edge Collapse Decimation: set a target face count and enable boundary preservation. It reduces density on flat areas while keeping detail where curvature demands it.
Does MeshLab support colors and textures?
Yes, within format limits. PLY preserves vertex colors and scan attributes, OBJ keeps UVs and materials, and photos can be projected onto geometry. Exporting to STL strips all color because the STL format itself only stores bare triangles.
What hardware does MeshLab need?
There are no official minimums, but community consensus suggests 8 GB of RAM for typical printing models and 32 GB for dense photogrammetry or LiDAR work, plus a GPU with OpenGL 3.0 or newer for a smooth viewport.