---
id: mkl-include
version: "2026.1.0"
license: Intel Simplified Software License
license_treatment: unclear
maintenance: active
---

# mkl-include — Intel® oneAPI Math Kernel Library

License: unclear · Maintenance: active · Downloads: 172.5K/mo

## What it is and what it does

mkl-include is Intel's Python package providing C and DPC++ language bindings to the oneAPI Math Kernel Library, a collection of highly optimized mathematical routines for linear algebra, Fourier transforms, and related computations. It targets developers building performance-critical numerical applications on Intel hardware. The package ships platform-specific wheels for Linux x86_64 and Windows x64, and depends on onemkl-license at runtime. It is not a standalone library—it provides headers and interfaces to call Intel MKL from C or C++ code, typically used by other libraries or compiled extensions rather than directly from Python.

The package is actively maintained and has been in production use since 2017. A key gotcha is that the MKLROOT environment variable is not set automatically during PyPI installation and must be configured manually according to Intel's developer documentation. This makes it less of a drop-in dependency and more of a system integration step, particularly relevant for containerized or CI/CD deployments.

Use it for:

- Building compiled Python extensions (e.g., NumPy, SciPy) that link against Intel MKL for faster linear algebra
- Developing C/C++ applications that call MKL routines and need the header files and interface definitions
- Optimizing numerical workloads on Intel CPUs by providing access to threaded BLAS, LAPACK, and FFT kernels
- Setting up development environments where MKL-accelerated libraries are built from source

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Provides C and Data Parallel C++ (DPC++) interfaces to Intel oneAPI Math Kernel Library for optimized mathematical computations on Intel CPUs and GPUs.

Yes, if you are building or deploying MKL-dependent compiled extensions on Linux x86_64 or Windows x64 and are willing to manually configure MKLROOT. No, if you expect a pure-Python package or automatic environment setup. The proprietary license and platform limitations mean it is not a general-purpose math library—it is a system integration tool for Intel MKL users.

## Install

pip install mkl-include
uv add mkl-include
poetry add mkl-include

## Installing mkl-include

Before you install: Medium install friction due to platform-specific wheels (Linux x86_64 and Windows x64 only) and a runtime dependency on onemkl-license. The package is actively maintained with a recent release.

License in practice: Licensed under Intel Simplified Software License (proprietary), classified as unclear treatment. Users should review Intel's license terms before deploying in commercial or restricted environments.

Quickstart:

pip install mkl-include

# After installation, set MKLROOT environment variable manually
# (not set automatically by PyPI installation)
import ctypes
# Link against MKL C interface via ctypes or compiler flags

MKLROOT environment variable must be configured manually post-installation; platform limited to Linux x86_64 or Windows x64; requires onemkl-license package.

Verify before relying:

- Whether onemkl-license is freely available or requires separate licensing from Intel
- Specific Python version compatibility (requires_python is unspecified)
- Whether MKLROOT setup documentation covers all common deployment scenarios

## Package facts

- License: Intel Simplified Software License (unclear)
- Python support: unspecified
- Install friction: medium
- Maintenance: active
- Downloads: 172.5K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

intel mkl c interface, oneapi math kernel library, optimized linear algebra, intel cpu gpu math library, high-performance computing math, dpc++ math routines, threaded math kernels, hpc, compiled-dependency, intel-specific

[View on SkillFed](https://skillfed.io/packages/mkl-include) · [View on PyPI](https://pypi.org/project/mkl-include/)