SupaVector Developer

Memory API documentation and quickstart

Create a memory, connect knowledge, test retrieval, and integrate the /v1/memories APIs from your backend.

What this page covers

Covers memory creation, search, ask, boolean_ask, chat, write, sync, and source management.
Shows how to use the exact memory ID and a server-side token in production code.
Explains source sync, direct writes, retrieval testing, and grounded answer workflows.

Common questions

Can SupaVector work with LangChain?

Yes. SupaVector can sit behind LangChain or your own runtime as the retrieval and durable memory layer while your application keeps its own orchestration and UX.

What can I call from my backend?

Use one memory-scoped API for search, grounded ask, boolean ask, chat, writes, sync, and source operations.