Store your knowledge and memory
Organize source documents, embeddings, and metadata in collections. Choose index types for the fields your application needs to search or filter. Supply your own vectors, or use native embeddings to generate document and query vectors from text and images with provider credentials managed by LambdaDB.Search with the right signals
Find relevant documents using full-text, dense vector, or sparse vector search. Combine retrieval paths with hybrid search, then optionally apply native reranking to evaluate the retrieved text against your question. Use filters to narrow matches and keyword facets to show counts alongside supported lexical queries. The search overview covers query options and links to each search method’s detailed guide.Version and recover your data
Use branches, tags, and aliases to isolate changes, pin datasets for evaluation, and switch application reads between versions. Collections retain committed snapshots for 30 days by default, with configurable retention. Create a point-in-time branch from retained history to inspect or recover an earlier state.Get started
Connect your coding agent to LambdaDB and run your first search, or build directly with an SDK. For direct HTTP requests, see the REST API reference.Start with an AI agent
Connect your coding agent and run your first search.
Start with an SDK
Create a collection, add documents, and run your first search with an SDK.
Use the CLI
Install the CLI, import JSONL, and run your first query.
Use the MCP server
Configure an MCP client manually and troubleshoot connections.