SvectorDB

SvectorDB is a serverless vector database built for AWS, offering cost-effective vector search with pay-per-request pricing. It provides hybrid search using Lucene/ElasticSearch-style filters, instant upserts and deletions, and built-in vectorizers for text and images. Users can bring their own embeddings or use SvectorDB's embed API. The database supports CloudFormation integration and offers SDKs for JavaScript, Python, and OpenAPI. Use cases include recommendation engines, document/image search, and retrieval augmented generation (RAG). Pricing is transparent: $0.25/GB/month for storage, $5/million queries, and $20/million writes. A free tier allows up to 5k records per database. Known limitations include no snapshot support and a default max of 1 million records per database.