manticore-search

Manticore Search introduces automatic document chunking for vector search, improving recall for long documents by splitting them into chunks and embedding each chunk.

Manticore Search 27.1.5 introduces a new ONNX Runtime backend for embeddings that achieves ~14x faster performance than the previous SentenceTransformers/Candle path, with throughput ranging from 70-230 docs/sec compared to 5-11 docs/sec, and no API changes required.

Manticore Search introduces early termination for HNSW-based KNN vector search, reducing distance computations by up to 80% for large k values while maintaining precision within 2-4% of full search.