gtsam
Georgia Tech Smoothing And Mapping library
Decision gist · record as of 2026-08-14
Yes, if you are working on robotics or computer vision research involving SLAM, visual odometry, or sensor fusion. The library is production-stable, actively maintained, and has no known vulnerabilities. Install friction is moderate due to compiled bindings, but prebuilt wheels for modern Python versions on common platforms mitigate this. The permissive BSD license poses no restrictions. Not necessary for general-purpose optimization or machine learning; consider it only if your problem naturally fits the factor-graph abstraction.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires a C++ compiler and build tools during wheel compilation if installing from source; prebuilt wheels avoid this for supported platforms.
- Medium install friction due to compiled C++ bindings, but prebuilt wheels are available for modern Python versions (3.11–3.14) on macOS (x86_64 and arm64) and Linux (x86_64 and aarch64).
- Maintenance is active with a recent release.
License · maintenance · safety
Simplified BSD license (permissive) — Permissive Simplified BSD license allows commercial and private use with minimal restrictions, making it suitable for proprietary robotics and vision projects.
last release 2026-08-04 (10 days)
0 known vulnerabilities (OSV.dev, 2026-08-14) · 81,499 downloads/mo, #14,223 on PyPI
Alternatives
Verify before relying
pip install gtsam
import gtsam
import numpy as np
# Create a factor graph and add factors for optimization
graph = gtsam.NonlinearFactorGraph()
initial_estimate = gtsam.Values()- Whether the Python wrapper exposes all C++ library features or a subset thereof
- Performance characteristics and scaling limits for large-scale SLAM problems
- Compatibility with specific robotics frameworks (ROS, etc.) beyond the core library
What it is and what it does
GTSAM is a robotics and computer vision library centered on factor graphs—a graphical model representation that encodes optimization problems arising in simultaneous localization and mapping (SLAM), visual odometry, and sensor fusion. Rather than working directly with sparse matrices, GTSAM uses factor graphs and Bayes networks to express relationships between poses, landmarks, and measurements, then solves them via nonlinear optimization. The Python wrapper provides access to this C++ core, letting you build and solve factor graphs for problems like visual-inertial odometry (combining camera and IMU data), pose-graph optimization, and multi-robot localization.
The library includes specialized factors for IMU preintegration—a state-of-the-art scheme for fusing inertial measurements with visual or other pose constraints—and supports both batch and incremental solving. It is widely used in academic robotics research and production systems. Runtime dependencies are minimal (numpy and pyparsing), and prebuilt wheels are available for common platforms, though the compiled nature of the bindings means installation may require a C++ toolchain on some systems.
Use it for
- Build and solve pose graphs for visual SLAM systems that fuse camera and IMU measurements
- Implement visual-inertial odometry (VIO) by preintegrating IMU data between keyframes
- Optimize multi-robot localization by adding relative pose constraints to a shared factor graph
- Refine camera calibration and landmark positions in structure-from-motion pipelines
- Prototype robotics perception algorithms that require nonlinear least-squares optimization over manifolds
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you are working on robotics or computer vision research involving SLAM, visual odometry, or sensor fusion.
The library is production-stable, actively maintained, and has no known vulnerabilities. Install friction is moderate due to compiled bindings, but prebuilt wheels for modern Python versions on common platforms mitigate this. The permissive BSD license poses no restrictions. Not necessary for general-purpose optimization or machine learning; consider it only if your problem naturally fits the factor-graph abstraction.
Install
gtsam on PyPI
Before you install
Medium install friction due to compiled C++ bindings, but prebuilt wheels are available for modern Python versions (3.11–3.14) on macOS (x86_64 and arm64) and Linux (x86_64 and aarch64). Maintenance is active with a recent release.
Requires a C++ compiler and build tools during wheel compilation if installing from source; prebuilt wheels avoid this for supported platforms.
License in practice
Permissive Simplified BSD license allows commercial and private use with minimal restrictions, making it suitable for proprietary robotics and vision projects.
Quickstart
pip install gtsam
import gtsam
import numpy as np
# Create a factor graph and add factors for optimization
graph = gtsam.NonlinearFactorGraph()
initial_estimate = gtsam.Values()Verify before relying
- Whether the Python wrapper exposes all C++ library features or a subset thereof
- Performance characteristics and scaling limits for large-scale SLAM problems
- Compatibility with specific robotics frameworks (ROS, etc.) beyond the core library
Package facts
| License | Simplified BSD license permissive |
| Python support | Not specified |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | 2 packagesnumpypyparsing |
| Maintenance | Actively maintained 10 days since the last release |
| First released | |
| Downloads | 81,499 / month, #14,223 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 5 - Production/StableIntended Audience :: DevelopersIntended Audience :: EducationIntended Audience :: Science/ResearchLicense :: OSI Approved :: BSD LicenseOperating System :: MacOSOperating System :: Microsoft :: WindowsOperating System :: POSIXProgramming Language :: Python :: 2Programming Language :: Python :: 3 |
Evidence: gtsam-4.2.2-cp311-cp311-macosx_10_15_x86_64.whl; gtsam-4.2.2-cp311-cp311-macosx_11_0_arm64.whl; gtsam-4.2.2-cp311-cp311-manylinux2014_aarch64.manylinux_2_17_aarch64.whl; gtsam-4.2.2-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.whl; gtsam-4.2.2-cp312-cp312-macosx_10_15_x86_64.whl; gtsam-4.2.2-cp312-cp312-macosx_11_0_arm64.whl; gtsam-4.2.2-cp312-cp312-manylinux_2_28_aarch64.whl; gtsam-4.2.2-cp312-cp312-manylinux_2_28_x86_64.whl; gtsam-4.2.2-cp313-cp313-macosx_10_15_x86_64.whl; gtsam-4.2.2-cp313-cp313-macosx_11_0_arm64.whl; gtsam-4.2.2-cp313-cp313-manylinux_2_28_aarch64.whl; gtsam-4.2.2-cp313-cp313-manylinux_2_28_x86_64.whl; gtsam-4.2.2-cp314-cp314-macosx_10_15_x86_64.whl; gtsam-4.2.2-cp314-cp314-macosx_11_0_arm64.whl; gtsam-4.2.2-cp314-cp314-manylinux_2_28_aarch64.whl; gtsam-4.2.2-cp314-cp314-manylinux_2_28_x86_64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “factor graph optimization”
- gtsamGTSAM is a Python wrapper around a C++ library for smoothing and…
- riskfolio-libRiskfolio-Lib builds optimized investment portfolios using…
- jaxliejaxlie implements Lie groups (SO2, SE2, SO3, SE3) for rigid body…
Give your agent the search over MCP, or paste the wish link into any chat.
More Scientific/Engineering packages
NumPy provides an N-dimensional array object and a comprehensive suite of mathematical, linear algebra, Fourier transform, and random number functions for scientific computing in Python.
pandas provides fast, flexible data structures (Series and DataFrame) for loading, cleaning, transforming, and analyzing labeled or relational data in Python.
scipy provides numerical algorithms for mathematics, science, and engineering—including optimization, integration, linear algebra, Fourier transforms, signal and image processing, and ODE solvers—built on numpy arrays.
scikit-learn provides a comprehensive Python library for supervised and unsupervised machine learning, including classification, regression, clustering, dimensionality reduction, and model evaluation tools built on NumPy and SciPy.
Install it if you need to train, evaluate, or deploy supervised or unsupervised learning models.
dill extends Python's pickle module to serialize and deserialize a much wider range of Python objects, including functions, lambdas, classes, and interpreter sessions, to byte streams for storage or network transmission.
Multiprocess is an enhanced fork of Python's standard multiprocessing library that uses dill for better serialization, allowing you to spawn processes with a threading-like API and share complex objects between them.
Install it if you use multiprocessing and encounter pickle serialization limits with lambdas or complex objects.
See also toppra · rsl-rl-lib · pykalman · mjlab · dex-retargeting · umap-learn · onemkl-sycl-sparse · mkl · simdkalman · mplib