---
id: scipy-openblas32
version: "0.3.34.106.0"
license: BSD 2-Clause License Copyright (c) 2021, MacPython All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following… (full text in the JSON record)
license_treatment: permissive
maintenance: active
---

# scipy-openblas32 — Provides OpenBLAS for python packaging

License: permissive · Maintenance: active · Downloads: 410.9K/mo

## What it is and what it does

scipy-openblas32 is a wheel distribution of OpenBLAS (a high-performance BLAS/LAPACK library) designed to serve as a build and runtime dependency for NumPy and SciPy. It bundles the OpenBLAS shared library, LAPACK, and required GCC runtime support into a self-contained wheel, eliminating the need to install system BLAS libraries separately. The package provides Python functions to query the location of headers, libraries, and pkg-config metadata for integration into build systems.

The primary use case is as a vendored dependency in NumPy and SciPy wheels, or as a build-time dependency in CI and local development. While other projects may use it for the same purposes, the maintainers explicitly warn against adding it as a runtime dependency outside NumPy and SciPy, because the package makes no backwards compatibility guarantees for symbol names or the Python API, and future changes could cause symbol conflicts or breakage.

Use it for:

- Building NumPy or SciPy wheels with a consistent, pre-built OpenBLAS to avoid system library version mismatches.
- Local development of NumPy or SciPy with a pinned OpenBLAS version to match CI environments.
- Vendoring into NumPy or SciPy wheels to ensure end-users receive a working BLAS without external dependencies.
- Providing OpenBLAS headers and libraries to custom build systems (via get_include_dir, get_lib_dir, get_pkg_config) during package compilation.

## Worth the install?

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

Provides pre-built OpenBLAS shared libraries and headers for use as a build or runtime dependency in NumPy, SciPy, and other Python projects that need linear algebra acceleration.

Yes, but only if you are building or developing NumPy, SciPy, or a closely related project that needs a pre-built OpenBLAS. Do not add it as a runtime dependency to other projects—the maintainers explicitly warn against this due to lack of stability guarantees and risk of symbol conflicts. For general linear algebra needs, depend on NumPy or SciPy instead.

## Install

pip install scipy-openblas32
uv add scipy-openblas32
poetry add scipy-openblas32

## Installing scipy-openblas32

Before you install: Medium install friction due to platform-specific wheels across multiple architectures (x86_64, ARM, PowerPC, s390x, RISC-V, and Windows variants). Actively maintained with recent releases; last commit 2026-08-10 and latest release 2026-08-09.

License in practice: BSD 2-Clause (MacPython wrapper) with bundled OpenBLAS (BSD-3-Clause), LAPACK (BSD-3-Clause), GCC runtime (GPL-3.0-with-GCC-exception), and libquadmath (LGPL-2.1-or-later). Permissive overall but includes GPL-licensed components; review bundled licenses if distributing.

Quickstart:

pip install scipy-openblas32==0.3.34.106.0

import scipy_openblas32
include_dir = scipy_openblas32.get_include_dir()
lib_dir = scipy_openblas32.get_lib_dir()
pkg_config = scipy_openblas32.get_pkg_config()

Intended primarily for NumPy and SciPy build systems or CI/local development; the package explicitly warns against adding it as a runtime dependency outside those projects due to lack of backwards compatibility guarantees and risk of symbol conflicts.

Verify before relying:

- Whether importing scipy_openblas32 at runtime (outside build context) is safe for non-NumPy/SciPy projects given the symbol-conflict warning.
- Specific version pinning requirements beyond ==; the description recommends == pins but does not detail breaking changes between patch versions.
- Performance characteristics or tuning options for the bundled OpenBLAS build (e.g., threading, precision).

## Package facts

- License: BSD 2-Clause License Copyright (c) 2021, MacPython All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following… (full text in the JSON record) (permissive)
- Python support: supports_current
- Install friction: medium
- Maintenance: active
- Downloads: 410.9K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

openblas wheels python, blas lapack shared library, numpy scipy build dependency, linear algebra acceleration, openblas prebuilt binaries, build-dependency, blas-lapack, numpy-scipy

[View on SkillFed](https://skillfed.io/packages/scipy-openblas32) · [View on PyPI](https://pypi.org/project/scipy-openblas32/)