Weaviate is an open-source, cloud-native vector database that stores both objects and their vectors, so semantic search, keyword filtering and retrieval-augmented generation can happen in a single system. It combines vector similarity search with structured filtering, supports hybrid queries that mix vectors with BM25 keyword search, and can vectorize your data at import time using integrated models from providers such as OpenAI, Cohere and Hugging Face, or accept pre-computed embeddings you bring yourself. Weaviate has collected nearly 17,000 stars on GitHub and is actively developed in 2026.
Weaviate is built for production search workloads: RAG systems, semantic and image search, recommendation engines, chatbots and content classification. It is cloud-native by design, with sharding, replication, multi-tenancy and role-based access control, and it offers GraphQL and REST APIs plus client libraries for common languages. You can run it locally with Docker, self-host it on Kubernetes, or use the managed Weaviate Cloud. Modules for reranking and generative search let you assemble complete retrieval pipelines, asking questions over your data and generating answers from the retrieved passages without stitching together separate tools.
Weaviate itself is free and open source to self-host. Weaviate Cloud offers a managed, serverless deployment on paid plans, so the overall offering is best described as freemium.
Weaviate is for developers and platform teams building semantic search, RAG applications and recommendation systems who want an open-source vector database that scales from a laptop prototype to a replicated production cluster.