# ashvardanian/StringZilla

Up to 100x faster strings for C, C++, CUDA, Python, Rust, Swift, JS, & Go, leveraging NEON, AVX2, AVX-512, SVE, GPGPU, & SWAR to accelerate search, hashing, sorting, edit distances, sketches, and memory ops 🦖

Repository: https://github.com/ashvardanian/StringZilla
Canonical: https://ross.abutalabs.com/products/stringzilla
Homepage: https://ashvardanian.com/posts/stringzilla/
Language: C
License: Apache-2.0
License Family: permissive
Topics: simd, dataset, string, string-manipulation, string-matching, substring, information-retrieval, sorting-algorithms, string-parsing, string-search, parser, hash, hashing, levenshtein-distance, edit-distance, gpu, search, unicode
Last push: 2026-08-20T17:07:06+00:00

## Health v2 (maintenance only)
Score: 94/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 98, release rhythm 85, longevity 100
- inputs: {"age_days": 2210, "days_push": 13, "days_rel": 22, "gap_med": 2, "n_releases_24m": 70}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3542, forks 132 (observed 2026-08-28T04:08:09.289010+00:00)

## What it is
StringZilla is a high-performance string processing library for C, C++, Python, Rust, Swift, JS, and Go that uses SIMD, SWAR, and GPU instructions to accelerate substring search, hashing, sorting, edit distances, and memory operations. It claims 5-100x speedups over standard library implementations like LibC, ICU, and NVIDIA's own GPU libraries.

## Use cases
- speed up substring search in large text files
- compute Levenshtein edit distance faster
- parse multi-terabyte newline-delimited files in Python
- accelerate string hashing and sorting
- fuzzy string matching at scale
- UTF-8 segmentation and tokenization faster than ICU
- GPU-accelerated sequence alignment with NW and SW algorithms

## When to choose
- you need maximum string processing throughput beyond standard library performance
- you process very large text datasets in Python that overwhelm native str
- you need fast edit distances on CPU or GPU
- you want portable SIMD-accelerated string primitives across many languages

## When to avoid
- your string workloads are small and not performance-critical
- you need a full-featured Unicode text processing framework with rich locale support
- you prefer staying purely within standard library APIs for maintainability

## Facets
- artifact type: library
- maturity: active
- function: parser, search-engine, serialization, benchmarking, gpu-computing
- domain: developer-tools, performance
- platform: cross-platform, python, rust, cpp, c, cli
- tags: simd, swar, string-processing, edit-distance, substring-search, hashing, unicode, memory-mapped-files, algorithms, natural-language-processing, data-engineering, gpu

## Member repositories
- ashvardanian/StringZilla (main) score 94

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:09.289010+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 2026-08-29T18:34:27.738682+00:00, confidence not recorded.
  - readme: https://github.com/ashvardanian/StringZilla (fetched 2026-08-28T04:08:09.289010+00:00, sha 5c1c1a36e1f9)
  - homepage: https://ashvardanian.com/posts/stringzilla/ (fetched 2026-08-29T09:28:33.173240+00:00, sha 4b59c0afd2de)
- Data as of 2026-08-30T08:39:29.467469+00:00.