Introduction

Welcome to anansi! This documentation should get you up and running quickly while learning about the core features that you can leverage. If you are looking for the HTTP1 api spec, you might want to start here.

What is anansi?

anansi is a standalone service and embeddable library that makes it easy for you to quickly add semantic search and LLM memory to your applications, without needing to rely on Python. More specifically, anansi allows developers to:

• Generate embeddings for text or images using state-of-the-art models on the Massive Text Embedding Leaderboard.
• Compare embeddings using various metric spaces: (Cosine, Hamming, L2, L1).
• Fine tune their embeddings via a live continuous process (WIP).

In the coming weeks (Q3 2023) anansi will be providing a managed service to further reduce adoption and setup costs.

Getting Started

Please refer to the detailed documenation for:

• Embedds (embedding generation)
• Horus (embedding search, WIP)

Contributing

We welcome contributions of all sizes and contributors at all levels! Please take a look at open issues on GitHub or hop into the Discord and take a look at #contributions.