Download
The GDAL project distributes GDAL as source code and Containers only. Binaries produced by others are available for a variety of platforms and package managers.
Source Code
Current Release
2024-08-16 gdal-3.9.2.tar.gz 3.9.2 Release Notes (3.9.2 md5)
Past Releases
Links to Past Releases are also available.
Development Source
The main repository for GDAL is located on GitHub at https://github.com/OSGeo/GDAL.
You can obtain a copy of the active source code by issuing the following command
git clone https://github.com/OSGeo/GDAL.git
Additional information is available about Build requirements and Building GDAL from source.
Maintenance policy
The GDAL upstream team only maintains the branch on which the latest release has been done, with bugfixes releases issued roughly every 2 months. So, for example, during the development phase of GDAL 3.10.0, GDAL 3.9.x bugfixes releases are done based on the release/3.9 branch, but not older branches (GDAL 3.8.x or older).
Binaries
In this section we list a number of the binary distributions of GDAL all of which should have fully reproducible open source build recipes.
Note that the maintainers of those distributions are generally not the maintainers of the GDAL sources, so please report any issue specific to those builds through their own support channels.
Windows
Windows builds are available via Conda Forge (64-bit only). See the Conda section for more detailed information. GDAL is also distributed by GISInternals and OSGeo4W and through the NuGet and vcpkg package managers.
Linux
Packages are available for Debian, Alpine_, Fedora_, and other distributions.
Mac OS
GDAL packages are available on Homebrew.
Android
GDAL can be installed using vcpkg. You may also refer to vcpkg Android support for general instructions.
For example to install default configuration for the arm64-android
target:
git clone https://github.com/Microsoft/vcpkg.git
cd vcpkg
./bootstrap-vcpkg.sh # ./bootstrap-vcpkg.bat for Windows
./vcpkg integrate install
export ANDROID_NDK_HOME=/path/to/android_ndk_home # to adapt
./vcpkg search gdal --featurepackages # list optional features
./vcpkg install gdal:arm64-android # install with default configuration
./vcpkg install gdal[poppler,netcdf]:arm64-android # install with Poppler and netdf support
Cross-Platform Package Managers
Conda
Conda can be used on multiple platforms (Windows, macOS, and Linux) to install software packages and manage environments. Conda packages for GDAL are available through conda-forge.
conda install [-c channel] [package...]
GDAL is available as several subpackages:
gdal
: Python bindings and Python utilities (depends on libgdal-core)libgdal
: meta-package gathering all below libgdal-* packages (except libgdal-arrow-parquet)libgdal-arrow-parquet
: (Geo)Arrow IPC File Format / Stream and (Geo)Parquet drivers as a plugin (depends on libgdal-core)libgdal-core
: core library and C++ utilities, with a number of builtin drivers (available since GDAL 3.9.1)libgdal-fits
: FITS -- Flexible Image Transport System driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-grib
: GRIB -- WMO General Regularly-distributed Information in Binary form driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-hdf4
: HDF4 -- Hierarchical Data Format Release 4 (HDF4) driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-hdf5
: HDF5 -- Hierarchical Data Format Release 5 (HDF5) driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-jp2openjpeg
: JP2OpenJPEG -- JPEG2000 driver based on OpenJPEG library driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-kea
: KEA driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-netcdf
: NetCDF: Network Common Data Form driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-pdf
: PDF -- Geospatial PDF driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-postgisraster
: PostGISRaster -- PostGIS Raster driver driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-pg
: PostgreSQL / PostGIS driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-tiledb
: TileDB - TileDB raster driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)libgdal-xls
: XLS - MS Excel format driver as a plugin (depends on libgdal-core, available since GDAL 3.9.1)
To install the gdal
package (Python bindings and utilities), and libgdal-core
:
conda install -c conda-forge gdal
To install the libgdal
meta-package with all available drivers, but libgdal-arrow-parquet:
conda install -c conda-forge libgdal
To install the Arrow and Parquet drivers as plugins:
conda install -c conda-forge libgdal-arrow-parquet
GDAL master Conda builds
GDAL master builds are available in the gdal-master
channel. They are based on dependencies from the conda-forge
channel.
First, install mamba into the base
environment, create a dedicated gdal_master_env
environment, and then activate the dedicated gdal_master_env
environment.
conda update -n base -c conda-forge conda
conda install -n base --override-channels -c conda-forge mamba 'python_abi=*=*cp*'
conda create --name gdal_master_env
conda activate gdal_master_env
Then install GDAL from the gdal-master
channel:
mamba install -c gdal-master gdal
mamba install -c gdal-master libgdal-arrow-parquet # if you need the Arrow and Parquet drivers
vcpkg
The GDAL port in the vcpkg dependency manager is kept up to date by Microsoft team members and community contributors. You can download and install GDAL using the vcpkg as follows:
git clone https://github.com/Microsoft/vcpkg.git
cd vcpkg
./bootstrap-vcpkg.sh # ./bootstrap-vcpkg.bat for Windows
./vcpkg integrate install
./vcpkg search gdal --featurepackages # list optional features
./vcpkg install gdal # install with default configuration
./vcpkg install gdal[poppler,netcdf] # install with Poppler and netdf support
If the version is out of date, please create an issue or pull request on the vcpkg repository.
Spack
Spack is a package management tool designed to support multiple versions and configurations of software on a wide variety of platforms and environments. It was designed for large supercomputing centers. Spack builds packages from sources, and allows tweaking their configurations.
You can find information about GDAL in Spack at https://packages.spack.io/package.html?name=gdal
For the default GDAL build with a reduced number of drivers:
git clone -c feature.manyFiles=true https://github.com/spack/spack.git
cd spack/bin
./spack install gdal
For a build with netcdf driver enabled:
./spack install gdal +netcdf
Containers
Docker images with nightly builds of GDAL master and tagged releases are available at GitHub Container registry.
Information on the content of the different configurations can be found at https://github.com/OSGeo/gdal/tree/master/docker.
Documentation
Besides being included when downloading the software, the documentation is also available independently as a PDF file, and a ZIP of individual HTML pages for offline browsing. (The ZIP also includes the PDF.) The documentation reflects the latest state of the development branch of the software.