mxnet
Apache MXNet is an ultra-scalable deep learning framework. This version uses openblas and MKLDNN.
Decision gist · record as of 2026-08-14
No, not for new projects. MXNet is abandoned—the repository is archived, the last release was in May 2022, and no security or maintenance updates are forthcoming. For new deep learning work, use an actively maintained framework. Install MXNet only if you are maintaining or migrating legacy code that already depends on it, and plan a migration path to a supported alternative.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- On Linux, libquadmath.so.0 must be installed separately: `sudo apt install libquadmath0` (Debian/Ubuntu) or `sudo yum install libquadmath` (RHEL/CentOS).
- Requires numpy and requests as runtime dependencies.
- Medium install friction due to compiled dependencies.
License · maintenance · safety
Apache 2.0 (permissive) — Licensed under Apache 2.0 (permissive), allowing commercial and private use with minimal restrictions. You must include a copy of the license and note any modifications, but there are no copyleft obligations.
last release 2022-05-17 (1550 days) · last repo commit 2023-10-25 · 20,811 stars · archived
0 known vulnerabilities (OSV.dev, 2026-08-14) · 693,094 downloads/mo, #5,318 on PyPI
Alternatives
Verify before relying
pip install mxnet
import mxnet as mx
import numpy as np
# Create a symbolic variable and layer
data = mx.sym.Variable('data')
fc = mx.sym.FullyConnected(data=data, num_hidden=10)- Whether the abandoned status and lack of recent updates pose compatibility risks with modern Python or dependency versions.
- Whether GPU-accelerated variants (mxnet-cu112, mxnet-cu110, etc.) remain functional or if CUDA support is also stalled.
- Current community adoption and whether alternative maintained frameworks are now preferred for new projects.
What it is and what it does
MXNet is a deep learning framework designed to balance efficiency and flexibility, allowing you to mix different programming styles and hardware targets in a single application. It provides symbolic and imperative programming interfaces, supports distributed training, and can run on CPUs, GPUs, and other accelerators. The package depends on numpy for numerical operations, requests for network communication, and graphviz for visualization.
The framework is mature and was widely used in production, but the repository is now archived and abandoned. The last release was in May 2022, over 1550 days ago. While the package itself may continue to work for existing projects, it receives no new features, bug fixes, or security updates. The codebase supports Python 3.5 through 3.8 according to its classifiers, and installation requires a system-level shared library on Linux.
Use it for
- Training neural networks for image classification or computer vision tasks using GPU acceleration.
- Building distributed deep learning pipelines across multiple machines or GPUs for large-scale model training.
- Deploying pre-trained models for inference in production environments where MXNet is already established.
- Prototyping neural network architectures using symbolic computation graphs for research or experimentation.
- Running legacy MXNet code or migrating existing projects that were built on this framework.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
No, not for new projects.
MXNet is abandoned—the repository is archived, the last release was in May 2022, and no security or maintenance updates are forthcoming. For new deep learning work, use an actively maintained framework. Install MXNet only if you are maintaining or migrating legacy code that already depends on it, and plan a migration path to a supported alternative.
Install
mxnet on PyPI
Before you install
Medium install friction due to compiled dependencies. The package requires libquadmath.so.0 on Linux systems (installed separately via apt or yum), and pre-built wheels are available for macOS and Linux x86_64/aarch64. Maintenance status is abandoned—the repository was archived after the last commit on 2023-10-25, over 1550 days ago.
On Linux, libquadmath.so.0 must be installed separately: `sudo apt install libquadmath0` (Debian/Ubuntu) or `sudo yum install libquadmath` (RHEL/CentOS). Requires numpy and requests as runtime dependencies.
License in practice
Licensed under Apache 2.0 (permissive), allowing commercial and private use with minimal restrictions. You must include a copy of the license and note any modifications, but there are no copyleft obligations.
Quickstart
pip install mxnet
import mxnet as mx
import numpy as np
# Create a symbolic variable and layer
data = mx.sym.Variable('data')
fc = mx.sym.FullyConnected(data=data, num_hidden=10)Verify before relying
- Whether the abandoned status and lack of recent updates pose compatibility risks with modern Python or dependency versions.
- Whether GPU-accelerated variants (mxnet-cu112, mxnet-cu110, etc.) remain functional or if CUDA support is also stalled.
- Current community adoption and whether alternative maintained frameworks are now preferred for new projects.
Package facts
| License | Apache 2.0 permissive |
| Python support | Not specified |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | 3 packagesnumpyrequestsgraphviz |
| Maintenance | Abandoned 1,550 days since the last release |
| Last repo commit | repository archived |
| First released | |
| Downloads | 693,094 / month, #5,318 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 :: Apache Software LicenseProgramming Language :: C++Programming Language :: CythonProgramming Language :: OtherProgramming Language :: PerlProgramming Language :: PythonProgramming Language :: Python :: 3.5Programming Language :: Python :: 3.6Programming Language :: Python :: 3.7Programming Language :: Python :: 3.8Programming Language :: Python :: Implementation :: CPythonTopic :: Scientific/EngineeringTopic :: Scientific/Engineering :: Artificial IntelligenceTopic :: Scientific/Engineering :: MathematicsTopic :: Software DevelopmentTopic :: Software Development :: LibrariesTopic :: Software Development :: Libraries :: Python Modules |
Evidence: mxnet-1.9.1-py3-none-macosx_10_13_x86_64.whl; mxnet-1.9.1-py3-none-manylinux2014_aarch64.whl; mxnet-1.9.1-py3-none-manylinux2014_x86_64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
An agent finds packages by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language. Give your agent the search over MCP.
More Software Development packages
Provides backported and experimental type hints for Python 3.9+, allowing use of newer typing features on older Python versions and enabling early experimentation with type system PEPs before they enter the standard library.
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.
FastAPI is a Python web framework for building REST APIs using type hints, with automatic request validation, serialization, and interactive API documentation.
Provides a way to document function parameters, class attributes, return types, and variables inline using Python's `Annotated` type hint syntax instead of traditional docstrings.
Typer builds command-line applications from Python functions using type hints, automatically generating help text, argument parsing, and shell completion.
Install it if you are building CLIs in Python.
Distlib provides low-level packaging utilities for building, distributing, and managing Python software—including metadata handling, version specifiers, wheel support, script installation, and dependency resolution.
See also multi-model-server · nvidia-cudnn-cu12 · model-archiver · nvidia-cudnn-cu13 · thinc · sagemaker-train · torch · nvidia-cudnn-cu11 · tensorflow · tensorflow-cpu