hora

Source Code

Hora is an efficient, open-source library for approximate nearest neighbor (ANN) search, written in Rust. It is designed for high-performance vector search, suitable for AI and machine learning applications.

Features

• High Performance: Written in Rust for speed and reliability, with performance comparable to C++ implementations.
• Approximate Nearest Neighbor Search: Implements a collection of algorithms for efficient ANN searches.
• Multi-language Support: Usable from Rust, Python, JavaScript (via WebAssembly), and Java.
• Multiple Index Types: Supports various index structures for different ANN scenarios.
• Multiple Distance Metrics: Supports several distance calculations for flexible similarity search.
• Portability: Designed to be portable across different environments.
• Reliability: Built with safe Rust, ensuring memory safety and robustness.
• Productivity: Offers high-level abstractions for ease of use.
• Examples and Demos: Provides example projects and online demos (e.g., face-matching, wine comment search).
• Open Source: Licensed under Apache License.

Installation

• Available via Cargo for Rust projects.
• Usable from Python, JavaScript (WebAssembly), and Java with provided bindings.

Pricing

Hora is open-source and free to use under the Apache License.