narwhals

Extremely lightweight compatibility layer between dataframe libraries

narwhals Permissive license MIT Active ★ 1,696 v2.24.0 released 2026-07-13

Install

narwhals on PyPI

pip

pip install narwhals

uv

uv add narwhals

poetry

poetry add narwhals

Package facts

License	MIT (permissive)
Python support	supports the current Python release (>=3.10)
Install friction	low — pure-Python wheel
Runtime dependencies	none
Maintenance	actively maintained — 31 days since the last release
Last repo commit	2026-08-13
First released	2024-02-20
Popularity	one of the top 1,000 most-downloaded packages on PyPI (30-day window, as of 2026-08-13)
Known vulnerabilities	none known (OSV.dev, checked 2026-08-13)

Evidence: narwhals-2.24.0-py3-none-any.whl

Keywords: dataframes, interoperability, pandas, polars, pyarrow, dask, modin, cudf

Development Status :: 5 - Production/StableOperating System :: OS IndependentProgramming Language :: Python :: 3Programming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14Programming Language :: Python :: Free Threading :: 2 - BetaTyping :: Typed

About narwhals

from the package's own PyPI description — quoted content, verbatim

Narwhals

PyPI version (image) Downloads (image) Trusted publishing (image) PYPI - Types (image) LFX Health Score (image) OpenSSF Scorecard (image)

Extremely lightweight and extensible compatibility layer between dataframe libraries!

Full API support: cuDF, Modin, pandas, Polars, PyArrow.
Lazy-only support: Daft, Dask, DuckDB, Ibis, PySpark,...

Read as markdown · JSON record · Source repository · Homepage · Docs

AI interpretation — verify before relying

AI-generated interpretation of the package facts above; every digit, version, license, or vulnerability id it cites is grounded in the facts already shown on this page

Narwhals is a zero-dependency compatibility layer that lets you write dataframe code once and run it against pandas, Polars, DuckDB, Dask, PySpark, cuDF, Modin, PyArrow, and other libraries without importing them.

Narwhals installs with zero runtime dependencies and sees active maintenance with a release 31 days ago. The pure-Python wheel distribution and support for Python 3.10–3.14 make installation frictionless across environments.

MIT license permits unrestricted commercial and private use, modification, and redistribution with minimal obligations—ideal for both library and application integration.

Usage

pip install narwhals

import narwhals as nw

def process_df(df):
    return nw.from_native(df).select('column').to_native()

# Works with pandas, Polars, DuckDB, etc.

Requires Python ≥3.10; the underlying dataframe library (pandas, Polars, etc.) must be installed separately by the user.

Verdict: Narwhals is a production-ready (Development Status 5), actively maintained library with no security vulnerabilities, zero dependencies, and full type hints. It solves a genuine pain point for library maintainers needing dataframe agnosticism without bloating their dependency tree. The top-1000 popularity tier and adoption by major projects (altair, bokeh, lightgbm, pandera) validate its utility.

Needs verification

Whether the stated 'negligible overhead' claim holds for your specific workload patterns and dataframe sizes.
Completeness of Polars API coverage for your particular use case—only a subset is supported.
Performance characteristics when chaining multiple lazy operations across different backends.

dataframe abstraction layerpolars-compatible pandas wrappermulti-dataframe library supportdataframe interoperabilitywrite once run anywhere dataframespandas polars duckdb compatibilitylazy eager dataframe api