ONNX (Open Neural Network Exchange)
Version 1.23.1
Maintained by ONNX project (hosted by the LF AI & Data Foundation)634
ONNX is an open format for representing machine learning models, covering both deep learning and traditional ML. It defines an extensible computation graph model, built-in operators, and standard data types, with a current focus on inference. Facebook and Microsoft started the ONNX community in 2017, and LF AI announced it as a graduate-level project in November 2019.17
- Website: Website (external site: onnx.ai)
- Repository: GitHub repository (external site: github.com)
- Documentation: Python package documentation (external site: onnx.ai)
- Documentation: Open governance (external site: github.com)
- License: License (Apache 2.0) (external site: raw.githubusercontent.com)
- Release notes: PyPI package (external site: pypi.org)
Availability and license
Overall availability
Documented as available to the general public. Access conditions and license terms may still apply.12
Availability is separate from permission: read the license before using or redistributing.
Source files carry the SPDX identifier Apache-2.0 with copyright attributed to the ONNX Project Contributors. The README points to the Linux Foundation's trademark usage rules for the ONNX mark.1
Public materials checklist
| Item | Status | Notes and evidence |
|---|---|---|
| Source codeIs the source code publicly readable? | Public | Developed in the public onnx/onnx GitHub repository.1 |
| DocumentationIs user documentation published? | Public | The repository links the IR specification, operator documentation, versioning principles, and Python API documentation.1 |
| InstallationAre installation instructions or packages publicly available? | Public | Installable from PyPI with pip install onnx; weekly builds are published as onnx-weekly.14 |
| Supported platformsAre supported operating systems or hardware documented? | Public | Version 1.23.1 requires Python 3.10 or later and ships wheels for Linux (x86_64 and aarch64), macOS (universal2), and Windows (x86, x64, and ARM64), plus a WebAssembly (Emscripten) wheel and a source distribution.45 |
| Release statusAre versioned releases published? | Public | Version 1.23.1 was released on PyPI on September 29, 2026.4 |
What it is useful for
Exchanging trained models between frameworks, tools, and hardware runtimes; the repository also provides a Python package with shape and type inference, opset version conversion, and a reference implementation.1
Organization context
U.S. eligibility
Eligible · basis: U.S.-governed project
ONNX is a hosted project of the LF AI & Data Foundation, a directed fund of The Linux Foundation; its PyPI contact list is on the LF AI & Data mailing-list domain and its README points to Linux Foundation trademark rules. The Linux Foundation is a 501(c)(6) nonprofit whose legal postal address is in San Francisco, California. Under the project's open governance, a five-member Steering Committee elected by community vote sets project direction, governance policy, and releases, with no more than one representative from any single member company; Special Interest Groups are responsible for specific parts of the project.68419103
Sources
This listing is not an endorsement, a safety assessment, or a federal approval.