This repository is the public home for all technical evaluations, experiments, and real-world demonstrations of Similarity API.

We use this space to openly publish reproducible benchmarks, accuracy studies, performance comparisons, and practical examples that showcase where Similarity API is faster or simpler than local approaches — including fuzzy matching libraries, embeddings-based similarity, custom dedupe pipelines, and rule-based systems.

Our philosophy is simple:

Prove the value of Similarity API in the open, using transparent and repeatable code — not marketing claims.

Every script in this repository is designed to be run as-is by any developer. You can validate the results yourself, plug in your own data, or adapt the examples directly into your own pipelines.

This repo will continue to grow.

For full API Documentaion, please visit https://www.similarity-api.com/documentation. For a free API key, please visit https://www.similarity-api.com/auth?tab=signup.