Alternatives to Arcana

pgvector adds vector similarity search inside Postgres, so embeddings live next to your relational data.

Meilisearch delivers typo-tolerant, sub-50ms search-as-you-type plus hybrid vector and multimodal retrieval through one REST API.

Vespa is an open-source AI search platform combining vector, text and tensor retrieval with machine-learned ranking in one engine.

PostgresML brings LLM inference, embeddings, and vector search to your database with SQL — Postgres with GPUs for RAG and ML.

VectorRAG.Net is an embedded .NET vector search library for in-process RAG, semantic search, and ANN retrieval.

TopK is a unified retrieval engine that runs multi-vector, dense, sparse, regex, and metadata-filtered search in one query, billed per unit of usage.

Fully managed vector database on Milvus for RAG, semantic search, and billion-scale AI applications.

Chroma is open-source search infrastructure for AI that keeps vectors, indexes and metadata on cheap object storage.

Fully managed vector database and knowledge platform: search, upsert, and rerank from one index and one API key.

Qdrant: open-source Rust vector search engine built for filtered hybrid search at scale

Vector and full-text search on object storage, with sub-10ms warm queries and pricing roughly 10x under compute-based vector databases.

TiDB is a distributed SQL database with native vector search for AI agent memory, state, and RAG at scale.

Weaviate is an open-source AI database that combines vector search, hybrid search, and RAG with agent memory in one stack.

LanceDB is an open-source multimodal lakehouse for vector search, curation, and training on one Lance table.

Lance is an open lakehouse format for multimodal AI, pairing a file format, table format, and catalog spec with hybrid vector and full-text search on object

Elasticsearch Labs: free tutorials, notebooks, and reference apps for AI search, vector search, and RAG on Elasticsearch.

Kronotop is a distributed transactional document database with built-in vector search and millions of isolated agent namespaces per cluster.

NodeDB is a universal database engine fusing vector, graph, document, columnar, key-value, and full-text search in one SQL binary.

Embedbase is a single API that combines a semantic search vector store with LLM text generation.

AI-native graph database running vector similarity and OpenCypher traversal in one query for GraphRAG pipelines.

Open-source, in-memory text vector database for .NET apps that adds semantic search and RAG without provisioning a vector server.

VectorAI DB: run a local vector database for persistent AI agent memory on your own hardware, offline or air-gapped.

Context layer for AI agents: push nested JSON once, then query graph, vector, and analytics through a single API.

Moss delivers in-process semantic search with sub-10ms retrieval for voice agents, copilots and on-device AI.

Open-source vector database for GenAI apps, scaling from a pip install to tens of billions of vectors

Vector embedding API that chunks, embeds, and writes vectors into Pinecone for you.

MongoDB Atlas keeps your vector embeddings in the same collection as your operational data, so RAG skips the sync pipeline.

Open-source in-process vector database from Alibaba — install one Python package, query millions of vectors in seconds.

Research lab training SID-1, an agentic search model for retrieving context in AI systems.

Open long-context text embeddings plus agentic code-compliance and submittal review for AEC firms