MIT Open Source

Open Source

When you're working on unsolved problems, you end up building internal libraries just to move forward. We open source ours, so engineers everywhere can benefit.

Memchunk

The fastest chunking library available. Achieve up to 1TB/s throughput for your text processing needs.

GitHub

Catsu

A lightweight, unified embeddings client with a clean, consistent API across all providers. Includes built-in retry logic, usage tracking, cost monitoring, and an improved developer experience.

Docs GitHub Embedding Registry

Chonkie

A lightweight ingestion library for fast, efficient, and robust RAG pipelines. Clean your data, chunk it, generate embeddings, and connect to any vector database with ease.

Docs GitHub