numexpr
Fast numerical expression evaluator for NumPy
Decision gist · record as of 2026-08-14
Yes. NumExpr is actively maintained, has no known vulnerabilities, and offers genuine performance benefits for large-array numerical workloads. The MIT license imposes no restrictions. Install it if you regularly work with large NumPy arrays and need to squeeze performance from complex expressions; skip it for small arrays or simple operations where overhead outweighs gains.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires NumPy; best performance gains on large arrays that exceed L1 CPU cache size.
- Medium install friction due to compiled wheels for multiple platforms and Python versions.
- Active maintenance with a recent release (27 days ago) and 2532 repository stars indicate solid ongoing support.
License · maintenance · safety
MIT (permissive) — MIT license is permissive; you can use, modify, and distribute this package with minimal restrictions, making it suitable for both open-source and proprietary projects.
last release 2026-07-18 (27 days) · last repo commit 2026-08-10 · 2,532 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 16,838,156 downloads/mo, #1,137 on PyPI
Alternatives
Verify before relying
pip install numexpr
import numpy as np
import numexpr as ne
a = np.arange(1e6)
b = np.arange(1e6)
result = ne.evaluate("a * b - 4.1 * a > 2.5 * b")- Whether Intel MKL integration (mentioned in docs) is available in standard pip wheels or requires separate configuration.
- Actual performance gains on your specific hardware and expression complexity (documentation notes 0.95x to 4x range but results vary).
- Free-threading support status on CPython 3.13+ and any threading interaction caveats beyond what the description states.
What it is and what it does
NumExpr is a numerical expression evaluator that accelerates array computations by parsing expressions into its own virtual machine bytecode and executing them in chunks that fit in CPU cache. Instead of creating intermediate arrays for each operation (as NumPy does), it processes small chunks of data across all available CPU cores, reducing memory pressure and improving cache locality. This approach typically yields speedups between 0.95x and 4x for common expressions, with larger gains possible for complex mathematical operations.
The package works by taking string expressions like `'a*b-4.1*a > 2.5*b'` and evaluating them on NumPy arrays without materializing temporary results. It supports arithmetic, comparison, and transcendental functions (sin, arcsinh, etc.), and can also operate on string arrays. Performance benefits are most pronounced on large arrays; small arrays may see minimal or no improvement. The package optionally integrates with Intel's MKL for further acceleration of transcendental functions on Intel architectures.
Use it for
- Accelerate large-scale numerical computations in data analysis pipelines where memory bandwidth is a bottleneck.
- Evaluate complex mathematical expressions on multi-gigabyte arrays without exhausting available RAM.
- Speed up scientific simulations that repeatedly apply the same expression to different array chunks.
- Parallelize array operations across CPU cores for workloads that would otherwise be single-threaded in NumPy.
- Reduce memory footprint in memory-constrained environments by avoiding intermediate array allocation.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes.
NumExpr is actively maintained, has no known vulnerabilities, and offers genuine performance benefits for large-array numerical workloads. The MIT license imposes no restrictions. Install it if you regularly work with large NumPy arrays and need to squeeze performance from complex expressions; skip it for small arrays or simple operations where overhead outweighs gains.
Install
numexpr on PyPI
Before you install
Medium install friction due to compiled wheels for multiple platforms and Python versions. Active maintenance with a recent release (27 days ago) and 2532 repository stars indicate solid ongoing support.
Requires NumPy; best performance gains on large arrays that exceed L1 CPU cache size.
License in practice
MIT license is permissive; you can use, modify, and distribute this package with minimal restrictions, making it suitable for both open-source and proprietary projects.
Quickstart
pip install numexpr
import numpy as np
import numexpr as ne
a = np.arange(1e6)
b = np.arange(1e6)
result = ne.evaluate("a * b - 4.1 * a > 2.5 * b")Verify before relying
- Whether Intel MKL integration (mentioned in docs) is available in standard pip wheels or requires separate configuration.
- Actual performance gains on your specific hardware and expression complexity (documentation notes 0.95x to 4x range but results vary).
- Free-threading support status on CPython 3.13+ and any threading interaction caveats beyond what the description states.
Package facts
| License | MIT permissive |
| Python support | Supports the current Python release >=3.11 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | 1 packagenumpy |
| Maintenance | Actively maintained 27 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 16,838,156 / month, #1,137 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 6 - MatureIntended Audience :: DevelopersIntended Audience :: Information TechnologyIntended Audience :: Science/ResearchOperating System :: Microsoft :: WindowsOperating System :: UnixProgramming Language :: PythonProgramming Language :: Python :: 3Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14Topic :: Software Development :: Libraries :: Python Modules |
Evidence: numexpr-2.14.2-cp311-cp311-macosx_10_9_x86_64.whl; numexpr-2.14.2-cp311-cp311-macosx_11_0_arm64.whl; numexpr-2.14.2-cp311-cp311-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl; numexpr-2.14.2-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; numexpr-2.14.2-cp311-cp311-musllinux_1_2_aarch64.whl; numexpr-2.14.2-cp311-cp311-musllinux_1_2_x86_64.whl; numexpr-2.14.2-cp311-cp311-win32.whl; numexpr-2.14.2-cp311-cp311-win_amd64.whl; numexpr-2.14.2-cp311-cp311-win_arm64.whl; numexpr-2.14.2-cp312-cp312-macosx_10_13_x86_64.whl; numexpr-2.14.2-cp312-cp312-macosx_11_0_arm64.whl; numexpr-2.14.2-cp312-cp312-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl; numexpr-2.14.2-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; numexpr-2.14.2-cp312-cp312-musllinux_1_2_aarch64.whl; numexpr-2.14.2-cp312-cp312-musllinux_1_2_x86_64.whl; numexpr-2.14.2-cp312-cp312-win32.whl; numexpr-2.14.2-cp312-cp312-win_amd64.whl; numexpr-2.14.2-cp312-cp312-win_arm64.whl; numexpr-2.14.2-cp313-cp313-macosx_10_13_x86_64.whl; numexpr-2.14.2-cp313-cp313-macosx_11_0_arm64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “fast numpy array expressions”
- numexprNumExpr evaluates numerical expressions on NumPy arrays with…
- snuggsSnuggs evaluates s-expressions (Lisp-like syntax) that operate on…
- pytensorPyTensor is a Python library for defining, optimizing, and evaluating…
Give your agent the search over MCP, or paste the wish link into any chat.
More Python Modules packages
Converts domain names between Unicode and ASCII-compatible encoding (Punycode) according to IDNA 2008 and Unicode Technical Standard 46, with security validation and broader script coverage than the standard library.
Install it if you work with internationalized domain names, need to validate domains, or use HTTP clients that depend on it transitively.
Setuptools is a Python build backend and package management tool that handles building, distributing, and installing Python packages, including support for C/C++ extension modules.
PyYAML parses and emits YAML 1.1 data format, enabling serialization and deserialization of configuration files and Python objects to and from human-readable YAML text.
Pydantic validates Python data structures against type hints, coercing and checking input at runtime to ensure it matches a declared schema.
Provides reusable metadata objects for use with PEP-593 `typing.Annotated` to express common constraints like bounds, collection sizes, and predicates on types.
Install it if you use or build libraries that need to express type constraints in a standardized, inspectable way—or if you want to annotate your own types with…
Provides runtime tools to inspect and introspect Python type annotations, enabling programmatic examination of type hints at execution time.
See also Theano-PyMC · Theano · onemkl-sycl-sparse · snuggs · mkl · onemkl-sycl-blas · array-record · pytensor · array-api-strict · Bottleneck