# Numba

NumPy aware dynamic Python compiler using LLVM

Repository: https://github.com/numba/numba
Canonical: https://ross.abutalabs.com/products/numba
Homepage: https://numba.pydata.org/
Language: Python
License: BSD-2-Clause
License Family: permissive
Topics: python, numpy, llvm, compiler, cuda, parallel, numba
Last push: 2026-08-26T00:16:18+00:00
Link (homepage): https://numba.pydata.org/

## Health v2 (maintenance only)
Score: 94/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 85, longevity 100
- inputs: {"age_days": 5291, "days_push": 8, "days_rel": 22, "gap_med": 55.5, "n_releases_24m": 11}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 11129, forks 1315 (observed 2026-08-28T04:10:46.163757+00:00)

## What it is
Numba is an open-source NumPy-aware JIT compiler that translates a subset of Python and NumPy code into fast machine code using LLVM. It supports automatic parallelization, GPU (CUDA) acceleration, ufunc creation, and C callbacks with minimal code changes.

## Use cases
- speed up numerical Python functions to near C/Fortran performance
- parallelize loops across CPU cores with minimal code changes
- compile Python code to run on NVIDIA GPUs via CUDA
- accelerate NumPy array computations and create custom ufuncs
- optimize Monte Carlo simulations and scientific algorithms
- use fast Python kernels inside Jupyter notebooks or with Dask/Spark

## When to choose
- your workload is numerical/array-based Python or NumPy code that is too slow
- you want C-like speed without rewriting in C/C++ or leaving the interpreter
- you need easy CPU parallelization or GPU offloading of numeric loops

## When to avoid
- your code is heavily object-oriented, uses arbitrary Python objects, or relies on unsupported libraries
- you need to compile non-numeric general-purpose Python
- you depend on features outside Numba's supported Python/NumPy subset

## Facets
- artifact type: library
- maturity: stable
- function: compiler, machine-learning, gpu-computing, concurrency, math
- domain: machine-learning, data-science, gpu-computing, performance
- platform: python, cross-platform
- tags: jit-compiler, numpy, llvm, cuda, scientific-computing, parallelization, high-performance-computing, algorithms, gpu

## Member repositories
- numba/numba (main) score 94
- numba/llvmlite (backend) score 94

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:46.163757+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:16:46.509544+00:00, confidence not recorded.
  - readme: https://github.com/numba/numba (fetched 2026-08-28T04:10:46.163757+00:00, sha 8fe3b3206db2)
  - homepage: https://numba.pydata.org/ (fetched 2026-08-29T08:15:21.718943+00:00, sha eccddb4e7a8c)
- Data as of 2026-08-30T08:39:29.467469+00:00.