Back to directory

Antfly

Antfly is a search-and-inference database written in Zig that combines full-text, dense, sparse and late-interaction vector search with graph traversal in one engine, running the models that chunk, embed, rerank and extract data inside the process.

Antfly

Antfly is a search-and-inference database, written in Zig with no external dependencies, that carries full-text (BM25), dense vector, sparse vector and late-interaction vector indexes plus graph traversal over the same table. Embeddings, chunks, extracted entities and graph edges are generated automatically as data is written, because the models that chunk, embed, rerank, transcribe, OCR and extract run inside the database process itself, and built-in RAG agents tie retrieval and generation together.

Hybrid search fuses full-text, RaBitQ-compressed dense vectors, SPLADE sparse vectors and ColQwen2 late-interaction vectors using reciprocal rank or relative score fusion. Other features include cross-encoder reranking, stats and facet aggregations, shard-level ACID transactions, automatic document expiration, and Postgres change-data-capture mirroring over logical replication. The same engine runs as a single file, a single node with a hot standby, a multi-Raft cluster, or serverless over object storage, and a CLI binary covers the server, tables, queries, backups, auth and a model registry.

More in RAG

See all 540
  • ai·rete·rag

    ai-rete-rag.com

    A platform that pairs a Rete rule engine with retrieval-augmented generation, so deterministic rules decide an outcome and grounded retrieval explains the decision in plain language.

  • Xtriever

    github.com

    A hybrid retrieval engine for retrieval-augmented generation, written in Rust to run on-device on a phone or a laptop, combining lexical BM25 search, dense embedding search, reciprocal-rank fusion and cross-encoder re-ranking with no server.

  • FileFlux

    github.com

    A .NET library that turns PDF, DOCX, HWP and other documents into optimized chunks for RAG systems, through a five-stage pipeline built on high-performance Rust FFI readers.