Ingestion
Chonkie
An ingestion library for robust context engineering. Clean, chunk, embed, and store your data in a few lines of code.
MIT open source
Software for everyone
Software improves when everyone can inspect, share, and contribute to the tools they depend on. Our work started in the open source, and we intend to keep contributing.
● github.com/feyninc
Chunking engine
Memchunk
Chunk at the speed of memory. Up to 1TB/s throughput powered by SIMD delimiter search, lookup tables, and zero-copy bindings.
Embeddings client
Catsu
A lightweight unified embeddings client. Enjoy a consistent API across all providers with built-in retries, usage tracking, and cost monitoring.
Benchmark
MTCB
Massive Text Chunking Benchmark
Measure retrieval quality across 9 different domains. Test the context you give your agents.
Contributors welcome