Qdrant MCP
UnknownQdrant is an open-source vector database for similarity search, and its official MCP server turns it into semantic memory for agents: two tools, qdrant-store and qdrant-find, save pieces of information and retrieve the ones closest in meaning, with no manual queries. It runs over stdio, SSE and HTTP; a local instance needs no key, while Qdrant Cloud requires QDRANT_URL and QDRANT_API_KEY.
Why this server
- Just two tools — qdrant-store and qdrant-find — minimal complexity for the agent
- Ready-made semantic memory — the agent decides what to remember and when to search for it
- A local Qdrant instance needs no key at all — Qdrant Cloud is only needed for scale
- Supports stdio, SSE and streamable HTTP — connects to any MCP client
Usage examples
Prompt: Remember: client Ivanov prefers calls on Tuesdays after 3pm
Result: The agent called qdrant-store, saving the note as a vector in the memory collection — retrievable in any future conversation.
Prompt: What have we already discussed about this client's preferences?
Result: qdrant-find retrieved the earlier note about Tuesday calls by semantic similarity, even without an exact word match.
Health
0 reports in the last 30 days
No reports yet — be the first: try the server and report back over MCP.
Reviews
No reviews yet.