rensa
High-performance MinHash implementation in Rust with Python bindings - 40x faster than datasketch
Decision gist · record as of 2026-08-14
Yes. Rensa is actively maintained, has no known vulnerabilities, carries a permissive MIT license, and offers substantial speed gains over comparable Python libraries (datasketch, FastSketch) with near-identical results. Install friction is moderate due to compiled bindings, but prebuilt wheels cover common platforms. Use it if you need to deduplicate or find similar items in large datasets and speed matters.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python >= 3.8.
- Prebuilt wheels available for Linux, macOS, and Windows; other platforms may require building from source.
- Medium install friction due to compiled Rust bindings, but wheels are prebuilt for common platforms (Linux, macOS, Windows) and Python versions.
License · maintenance · safety
MIT (permissive) — MIT license is permissive; you can use, modify, and distribute this package with minimal restrictions.
last release 2026-07-23 (22 days) · last repo commit 2026-08-04 · 371 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 96,884 downloads/mo, #13,188 on PyPI
Alternatives
Verify before relying
pip install rensa
from rensa import RMinHash
m1 = RMinHash(num_perm=128, seed=42)
m1.update("the quick brown fox".split())
m2 = RMinHash(num_perm=128, seed=42)
m2.update("the quick brown cat".split())
print(m1.jaccard(m2))- Whether the 608.52x speedup vs datasketch holds across all use cases or only the reference benchmark conditions.
- Whether C-MinHash's formal variance proofs provide practical advantages for specific deduplication thresholds.
- Performance characteristics when working with non-ASCII or very large token sets.
What it is and what it does
Rensa is a Rust-based MinHash library that estimates Jaccard similarity between sets and identifies near-duplicates in large datasets. It implements two variants: R-MinHash (Rensa's own, optimized for speed) and C-MinHash (based on published research with formal variance bounds). The core algorithm applies k random hash functions to a set and keeps the minimum value from each; sets sharing many elements produce similar minimums, and the fraction of matching slots estimates the Jaccard index.
Rensa replaces the traditional modular reduction step with multiply-shift hashing, which is faster on modern CPUs and naturally produces 32-bit signatures (half the memory of standard implementations). It includes batch APIs for bulk operations, LSH (Locality Sensitive Hashing) indexing for efficient candidate retrieval, and a deduplicator for end-to-end workflows. The library has no runtime dependencies and supports Python 3.8 and later on CPython and PyPy.
Use it for
- Find and remove near-duplicate documents or SQL queries in large datasets using LSH indexing and batch deduplication.
- Estimate Jaccard similarity between tokenized text sets without computing exact set intersection.
- Build a searchable index of document signatures to quickly retrieve candidates above a similarity threshold.
- Deduplicate synthetic or generated datasets where exact matching is too strict but similarity thresholds are appropriate.
- Batch-process thousands of token sets into MinHash signatures with minimal Python call overhead.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes.
Rensa is actively maintained, has no known vulnerabilities, carries a permissive MIT license, and offers substantial speed gains over comparable Python libraries (datasketch, FastSketch) with near-identical results. Install friction is moderate due to compiled bindings, but prebuilt wheels cover common platforms. Use it if you need to deduplicate or find similar items in large datasets and speed matters.
Install
rensa on PyPI
Before you install
Medium install friction due to compiled Rust bindings, but wheels are prebuilt for common platforms (Linux, macOS, Windows) and Python versions. Active maintenance with recent releases; last commit 2026-08-04.
Requires Python >= 3.8. Prebuilt wheels available for Linux, macOS, and Windows; other platforms may require building from source.
License in practice
MIT license is permissive; you can use, modify, and distribute this package with minimal restrictions.
Quickstart
pip install rensa
from rensa import RMinHash
m1 = RMinHash(num_perm=128, seed=42)
m1.update("the quick brown fox".split())
m2 = RMinHash(num_perm=128, seed=42)
m2.update("the quick brown cat".split())
print(m1.jaccard(m2))Verify before relying
- Whether the 608.52x speedup vs datasketch holds across all use cases or only the reference benchmark conditions.
- Whether C-MinHash's formal variance proofs provide practical advantages for specific deduplication thresholds.
- Performance characteristics when working with non-ASCII or very large token sets.
Package facts
| License | MIT permissive |
| Python support | Supports the current Python release >=3.8 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | None |
| Maintenance | Actively maintained 22 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 96,884 / month, #13,188 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Programming Language :: Python :: Implementation :: CPythonProgramming Language :: Python :: Implementation :: PyPyProgramming Language :: Rust |
Evidence: rensa-0.4.1-cp310-cp310-manylinux_2_17_i686.manylinux2014_i686.whl; rensa-0.4.1-cp310-cp310-manylinux_2_28_aarch64.whl; rensa-0.4.1-cp310-cp310-manylinux_2_28_armv7l.whl; rensa-0.4.1-cp310-cp310-manylinux_2_28_ppc64le.whl; rensa-0.4.1-cp310-cp310-manylinux_2_28_s390x.whl; rensa-0.4.1-cp310-cp310-manylinux_2_28_x86_64.whl; rensa-0.4.1-cp310-cp310-musllinux_1_2_aarch64.whl; rensa-0.4.1-cp310-cp310-musllinux_1_2_armv7l.whl; rensa-0.4.1-cp310-cp310-musllinux_1_2_i686.whl; rensa-0.4.1-cp310-cp310-musllinux_1_2_x86_64.whl; rensa-0.4.1-cp310-cp310-win_amd64.whl; rensa-0.4.1-cp311-cp311-macosx_10_12_x86_64.whl; rensa-0.4.1-cp311-cp311-macosx_11_0_arm64.whl; rensa-0.4.1-cp311-cp311-manylinux_2_17_i686.manylinux2014_i686.whl; rensa-0.4.1-cp311-cp311-manylinux_2_28_aarch64.whl; rensa-0.4.1-cp311-cp311-manylinux_2_28_armv7l.whl; rensa-0.4.1-cp311-cp311-manylinux_2_28_ppc64le.whl; rensa-0.4.1-cp311-cp311-manylinux_2_28_s390x.whl; rensa-0.4.1-cp311-cp311-manylinux_2_28_x86_64.whl; rensa-0.4.1-cp311-cp311-musllinux_1_2_aarch64.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 › “jaccard similarity fast”
- rensaRensa computes MinHash signatures for fast similarity estimation and…
- datasketchProvides probabilistic data structures (MinHash, HyperLogLog, and…
- simsimdSimSIMD provides SIMD-optimized kernels for computing vector…
Give your agent the search over MCP, or paste the wish link into any chat.
More Information Analysis packages
A drop-in replacement for Python's standard `re` module that adds advanced regex features like nested sets, fuzzy matching, lookaround in conditionals, and full Unicode case-folding while maintaining backward compatibility.
pyarrow provides Python bindings to Apache Arrow's C++ libraries for efficient columnar data processing, serialization, and interoperability with pandas, NumPy, and other Python ecosystem tools.
NetworkX provides data structures and algorithms for creating, analyzing, and manipulating graphs and networks, supporting everything from simple undirected graphs to complex directed and weighted networks.
Connects Python applications to Snowflake data warehouses using the DB API 2.0 specification, enabling SQL queries, data transfers, and warehouse operations.
ContourPy calculates contours of 2D quadrilateral grids using C++11 algorithms wrapped in Python, offering serial and multithreaded implementations without requiring Matplotlib as a dependency.
Snowpark Python provides APIs to query and process data directly in Snowflake without moving data to your local system, with support for both native Snowpark and pandas-compatible interfaces.
Install it if you use Snowflake and want to process data without moving it to your application layer.
See also datasketch · semhash · simhash · array-record · imagededup · mhfp · py-tlsh · murmurhash · xxhash · py-multihash