# Meshroom: Open-Source Node-Based Photogrammetry and 3D Reconstruction Software > Meshroom is a free visual programming toolbox built by the AliceVision team for constructing and executing complex image-processing pipelines, with a default plugin that turns photographs into 3D models. Its node graph engine caches intermediate results and invalidates only downstream nodes on attribute changes, making iterative reconstruction practical on ordinary workstations. **alicevision/Meshroom** — Node-based Visual Programming Toolbox - Repository: https://github.com/alicevision/Meshroom - Website: http://meshroom.org - Stars: 12,988 · Forks: 1,234 - Language: Python - License: NOASSERTION - Published: 2026-09-16 · Updated: 2026-09-16 · Language: en - Canonical page: https://hysenlabs.com/projects/alicevision-meshroom ## What Meshroom Is and Who It Is For Meshroom addresses a specific workflow problem: turning collections of photographs into 3D models, camera paths, or geometry outputs requires a chain of algorithms that are complex to wire together manually. Meshroom wraps this pipeline in a node-based visual editor, making it possible to inspect, modify, and re-run individual steps without reprocessing the entire dataset. The primary audience is photographers, researchers, and VFX professionals who need photogrammetry without a commercial tool subscription. The official releases bundle the AliceVision plugin, which provides production-grade 3D computer vision algorithms developed in collaboration between academic institutions and industry. Secondary users are developers who want to integrate photogrammetry into larger automated workflows, since Meshroom supports both a local GUI and distributed execution on render farms. ## The Node Graph Model: How Pipelines Work The central concept in Meshroom is the graph. A graph is a collection of interconnected nodes, where each node performs a specific task and output attributes feed into subsequent steps. The user builds a pipeline by connecting nodes in the Graph Editor. The invalidation system is the defining technical property. When a node's attribute is modified, only the nodes downstream of that change are marked as needing recomputation. Cached results for unaffected nodes are reused, which means changing a parameter in the middle of a multi-hour pipeline does not force the entire pipeline to restart. This makes the iterative tuning of photogrammetry parameters practical on hardware without a dedicated compute cluster. Templates are pre-built pipeline configurations provided by plugins. Users can start from a template, customise individual node attributes, and save the modified graph as a new template. The Node Editor shows multiple tabs for the selected node: Attributes for parameter control, Log for execution output, Statistics for resource consumption, Status for technical metadata, Documentation for the node's reference material, and Notes for user annotations. The 2D and 3D Viewer panels display output from specific nodes, and the Image Gallery shows the full list of input files. ## Installing Meshroom and Running a First Pipeline The fastest installation path is to download a pre-compiled binary from the GitHub releases page. Meshroom releases pre-built packages for supported platforms; no build environment is needed for ordinary use. Custom plugin development requires setting up the project from source following INSTALL.md, and plugin installation is documented separately in INSTALL_PLUGINS.md. For users building from source, the runtime dependencies listed in requirements.txt include psutil, pyseq, and PySide6. The PySide6 version is constrained by Python version: ``` PySide6==6.8.3; python_version >= "3.9" PySide6==6.5.3; python_version < "3.9" ``` PySide6 is the Qt6 binding for Python, which means Qt6 runtime libraries must be present on the system. The pre-compiled releases handle this packaging; source builds require platform-appropriate Qt6 setup. To start a reconstruction from photographs, open Meshroom and either use a bundled template from the AliceVision plugin or create a graph manually by placing nodes in the Graph Editor and connecting their attribute outputs to inputs. Set the input images on the FeatureExtraction node, then execute the graph locally. Meshroom's local and render-farm execution modes can run simultaneously. ## The AliceVision Plugin and Its Reconstruction Pipelines The default plugin bundled with official releases comes from the AliceVision project. AliceVision provides the core algorithms for six pipeline categories. The most common is 3D reconstruction from multi-view images, which converts a set of overlapping photographs into a point cloud, a mesh, and texture output. Camera tracking estimates camera motion across a sequence. HDR fusion combines multiple exposure brackets into a high dynamic range result. Panorama stitching assembles wide-field images, including fisheye lenses and motorised head setups. Photometric stereo reconstructs geometry from a single viewpoint lit by multiple light sources. Multi-view photometric stereo combines the wide-baseline coverage of photogrammetry with the surface-detail capture of photometric stereo. The AliceVision algorithms are described as research-grade with production-level robustness, developed through a collaboration between academic and industrial partners. The plugin results shown on Sketchfab and the plugin overview page at alicevision.org document the types of output Meshroom can produce with default settings. ## The Plugin Ecosystem Beyond AliceVision Meshroom's architecture supports third-party plugins, and the MeshroomHub organisation on GitHub hosts several maintained examples. mrSegmentation adds nodes for AI-powered image segmentation from natural language prompts, using foundation models to isolate objects or regions in images before reconstruction. mrDepthEstimation provides monocular depth estimation from image sequences using deep learning models. mrRoma integrates RoMa, a dense feature matching approach with per-pixel correspondence and certainty maps, which is useful for matching images under large viewpoint or illumination changes. mrGSplat adds 3D Gaussian Splatting reconstruction, connecting the AliceVision photogrammetry pipeline to a Gaussian splat output stage for new viewpoint rendering. Meshroom Research is a plugin for evaluating and benchmarking machine learning algorithms in 3D computer vision, providing experimental nodes and comparison frameworks for research use. One plugin carries a relevant limitation: MeshroomMicMac integrates the photogrammetric algorithms from MicMac, an open-source system from the French mapping agency IGN. The plugin does not yet support Meshroom's full invalidation system, so graph-level caching and selective re-execution do not apply to MicMac nodes. This is noted in the README as an ongoing gap. ## Where Meshroom Falls Short The most significant practical limitation is the AliceVision dependency. The core 3D reconstruction pipelines require GPU compute for certain stages, and the README for AliceVision documents specific hardware and driver requirements. Without a compatible NVIDIA GPU, GPU-accelerated stages fall back to CPU execution, which is substantially slower for large image sets. Meshroom does not provide a point-and-click simplified workflow for non-technical users. The node graph interface assumes familiarity with pipeline concepts: understanding which node's attribute to change when reconstruction quality is poor requires reading node documentation and understanding the algorithm chain. For teams that need a supported commercial product with professional technical support, Agisoft Metashape is a proprietary photogrammetry application with a desktop interface and commercial licensing. Metashape does not use a node-based pipeline model; it presents photogrammetry as a sequence of menu-driven processing steps, which is a different design philosophy that lowers the learning curve at the cost of the fine-grained pipeline control Meshroom provides. The repository lists no standard SPDX licence identifier. The top-level directory includes COPYING.md and LICENSE-MPL2.md, indicating at least some components are under MPL 2.0. Teams with legal requirements around open-source licence compliance should review COPYING.md and CONTRIBUTING.md carefully before incorporating Meshroom in a commercial pipeline. ## Maintenance and Licence Considerations The last push to the repository was on 2026-09-28. The most recent formal release was v2025.1.0, published 2025-08-19. A nightly build from the develop branch was published 2026-06-09. The gap between the formal release and the nightly suggests active ongoing development that has not yet been packaged into a stable release. The repository licence field shows NOASSERTION, meaning the project does not assert a single standard SPDX identifier. The presence of COPYING.md and a separate LICENSE-MPL2.md suggests the codebase has components under different terms. Before distributing Meshroom-based work or incorporating it into a commercial pipeline, review these files to understand which parts of the codebase fall under which terms. This is not a legal assessment; it is a flag that the licence situation requires independent verification. ## Conclusion Meshroom is the right choice for researchers, VFX artists, and engineers who need an open, extensible photogrammetry pipeline they can customise at the node level. It is not a good fit for users who want a point-and-click experience with no pipeline understanding required, or for production environments on platforms that have not been verified against the INSTALL.md and requirements.txt. Before adopting it, confirm that your GPU and OS combination is supported by the AliceVision libraries, and review the COPYING.md file for the project's licence terms, since the repository lists no standard SPDX identifier. ## FAQ ### Is Meshroom free to use? Meshroom is free and open-source. Pre-compiled binaries are available on the GitHub releases page at no cost. The repository's licence situation is not fully asserted under a single SPDX identifier, so users incorporating it in commercial pipelines should review COPYING.md before distributing. ### What is Meshroom software? Meshroom is an open-source visual programming toolbox for building and executing data processing pipelines, primarily used for photogrammetry and 3D reconstruction from photographs. It uses a node graph where each node represents one processing step, and output attributes flow into subsequent nodes. ### How do you install Meshroom? Download a pre-compiled binary from the GitHub releases page for the simplest installation. Users who need to build from source or integrate custom plugins should follow INSTALL.md and INSTALL_PLUGINS.md in the repository. The runtime requires PySide6 and psutil, as listed in requirements.txt. ### How do you use Meshroom for photogrammetry? Open Meshroom, load a template pipeline from the AliceVision plugin, set your input images on the relevant node, and execute the graph locally or on a render farm. The Graph Editor lets you modify individual node attributes and re-run only the affected downstream steps thanks to the invalidation caching system. ### Is Meshroom still a maintained project? The last push to the repository was on 2026-09-28, and a nightly build was published 2026-06-09. The most recent stable release was v2025.1.0 from 2025-08-19. ## Sources - [alicevision/Meshroom on GitHub](https://github.com/alicevision/Meshroom) - [Issues](https://github.com/alicevision/Meshroom/issues) - [Project website](http://meshroom.org) - [README](https://github.com/alicevision/Meshroom/blob/develop/README.md) - [Releases](https://github.com/alicevision/Meshroom/releases) --- Hysen Labs editorial analysis, written from the project's own repository and release notes. Cite the canonical page: https://hysenlabs.com/projects/alicevision-meshroom