Is the key scoped to one index?

Most of these servers take an account-level key and expose every collection on the instance. A key scoped to one index is the difference between a bad query and a wiped production index, and the vendor, not the server, is where you set that.

Chroma and Qdrant both run on your own machine. A local one keeps the vectors on disk, needs no key at all, and cannot leak an index to anyone. It is the right default for anything you would not publish.

Which write tools come with search?

Upsert and delete usually live in the same server as query. If you only need retrieval, a server that exposes only retrieval is a smaller thing to get wrong.

Topic

MCP servers for vector databases

Query and manage the store underneath your embeddings.

This is the storage layer a RAG setup sits on: Pinecone, Qdrant, Chroma, Weaviate, Milvus, or Postgres with pgvector. The servers here mostly wrap one vendor's client, so the thing that varies is not the search quality, it is how much of your instance a single key can reach.

Graded servers: 130
Grade A: 50
Run remotely: 34
Reviewed by users: 1

Before you install one

Three things that separate the servers here, none of which shows up in a description.

Is the key scoped to one index?: Most of these servers take an account-level key and expose every collection on the instance. A key scoped to one index is the difference between a bad query and a wiped production index, and the vendor, not the server, is where you set that.
Local or hosted?: Chroma and Qdrant both run on your own machine. A local one keeps the vectors on disk, needs no key at all, and cannot leak an index to anyone. It is the right default for anything you would not publish.
Which write tools come with search?: Upsert and delete usually live in the same server as query. If you only need retrieval, a server that exposes only retrieval is a smaller thing to get wrong.

Every listing below has a scan report saying what the server asks for and what it does with it. The rubric is on how we grade.

The vector database servers we would look at first

Reviewed servers rank first, then the ones with real adoption. At most two per publisher, so no account can fill the list. Nothing here is sponsored.

1
Qdrant MCP
by avaazquezz
B
The full Qdrant API as 58 MCP tools: plain-text semantic/BM25/hybrid search and a search advisor.
No reviews yet
local · pypi2677/wkFree
2
amu-pgvector
by sangaraju1988
B
Lineage-gated agent memory tools on Postgres + pgvector, enforced by row-level security.
No reviews yet
local · pypi2.6k/wkFree
3
Vector Store Scanner
by 4hmetuyar
A
Probes vector-database endpoints for unauthenticated exposure of embeddings/RAG data
No reviews yet
local · npm1369/wkFree
4
Pinecone
by mcparmory.com
B
Create and manage vector database indexes, backups, and collections
No reviews yet
local · mixed31149/wkFree
5
Figma Vector MCP
by workingmodel
A
Let Claude create and edit fully editable vector nodes on the Figma canvas.
No reviews yet
local · npm166/wkFree
6
Vector Store Scanner
by GuardBee
A
Probes vector-database endpoints for unauthenticated exposure of embeddings/RAG data
No reviews yet
local · npm1204/wkFree
7
GenPark Vector Embedding Cache Deduplication Sentinel
by alphaparkinc
A
High-dimensional vector embedding semantic SimHash deduplication sentinel with LRU cache simulation
No reviews yet
local · mcpb7Free
8
VectorCache Semantic Vector Cache
by pgalyen1987
A
x402-gated semantic vector cache for agent swarms: Redis-backed similarity lookup. Paid in USDC.
No reviews yet
local · npm24/wkFree
9
Weaviate
by mrfentmen
A
MCP server for Weaviate: collections, objects, GraphQL search. Works with any instance.
No reviews yet
local · npm84/wkFree
10
MLP Vector Club
by MLP-VectorClub
A
Search and edit the MLP Vector Club color guide: appearances, tags, shows and reservations.
No reviews yet
gateway-readyFree
11
Vector Knowledge Graph
by CSOAI-ORG
A
AI-powered vector knowledge graph MCP server for agents. Supports add node, add edge, semantic node
No reviews yet
local · pypi22/wkFree
12
Inkbolt vector and raster editor
by skulitom
A
Local vector and raster editing engine for agents: layered documents, saved sessions and exports.
No reviews yet
local · mcpb1Free

36 more, graded and ranked

13
Qdrant Memory
Persistent scoped Qdrant memory for agents, through four explicit stdio MCP tools.
not reviewed
New
B
14
Qdrant RBAC
Role-based access to your Qdrant vector store, via MCP. OIDC in, scoped JWTs out.
not reviewed
New
B
15
pgvector
Similarity search, hybrid search, and index management for pgvector-backed PostgreSQL tables
not reviewed
New
B
16
Milvus
MCP server for Milvus: collections, stats, search. Works with any instance.
not reviewed
New
A
17
Local Vector Search MCP
Local Markdown search, navigation and editing via MCP, with optional vector embeddings.
not reviewed
New
A
18
Vector Mirror
Deterministic SVG perception for LLM agents — measures, never guesses. MCP server.
not reviewed
New
B
19
Vector Memory
Semantic document memory using Redis vector store. Save and recall files with natural language.
not reviewed
New
B
20
Vector DB MCP
Assembly-accelerated DuckDB MCP server for code search, indexing, and symbol lookup.
not reviewed
New
B
21
Vector Panda
Operate your Vector Panda collections: search and upsert vectors, schema, measured indexes, exports.
not reviewed
New
B
22
Vector Panda Documentation
Search and read Vector Panda docs: API operations, pricing, storage tiers, measured benchmarks.
not reviewed
New
B
23
Codex Pets Registry
Search, preview, install community Codex pet packs, and discover the pet request flow.
not reviewed
New
B
24
Embeddings, Rerank & Speech-to-Text (Tanod)
Speech denoise, transcripts (SRT/VTT), embeddings, reranking, entities and translation.
not reviewed
New
B
25
Qdrant
Read-only Qdrant vector DB for AI agents — collections, info, browse points, count, search.
not reviewed
New
C
26
CogniCore
Memory, reflection and safety for AI agents. Hosted or local, no vector database.
5.0(1)
Growing
A
27
Appwrite
Build and deploy backends: auth, Postgres/MySQL/vector databases, storage, functions, hosting, CDN
not reviewed
Established
A
28
Alibaba Cloud DashScope — Qwen
Alibaba Cloud DashScope MCP: Qwen chat, embeddings, and model discovery on your own account.
not reviewed
Established
B
29
MCP AutoMem
Graph-vector memory for AI assistants using FalkorDB and Qdrant
not reviewed
Established
C
30
IBM watsonx.ai Granite
IBM watsonx.ai MCP by three.ws: chat, text generation, embeddings, and tokenization.
not reviewed
Established
B
31
Memtomem
Markdown-first long-term memory for AI agents. Hybrid BM25 + vector search, runs locally.
not reviewed
Growing
A
32
ArcadeDB MCP Server
Built-in MCP server for ArcadeDB multi-model database (graph, document, vector, time-series)
not reviewed
Growing
B
33
Svg Tracer
MCP server for converting bitmap images to SVG vector graphics using potrace tracing.
not reviewed
Growing
A
34
Srclight
Deep code indexing for AI agents. FTS5 + embeddings + call graphs + git intelligence. Fully local.
not reviewed
Growing
A
35
jevreel
Search and understand videos without embeddings: contact sheets plus Jev typed decisions.
not reviewed
Growing
B
36
Openai
MCP server for OpenAI API (chat completions, image generation, embeddings) via AceDataCloud
not reviewed
Growing
A
37
Local Memory
Persistent local AI memory. SQLite + FTS5 + vector search, knowledge graph. No cloud, no API keys.
not reviewed
Growing
A
38
Vecline
Local-first image, vector and document toolkit that measures its own output. Nothing uploaded.
not reviewed
Growing
A
39
Clude
Portable agent memory anchored on Solana. Local SQLite + vector recall, open export.
not reviewed
Growing
B
40
IT Glue
IT Glue MCP server for MSPs — documents, flexible assets, semantic vector search, RBAC, and BYOK
not reviewed
Growing
B
41
SAME - Stateless Agent Memory Engine
Persistent memory for AI coding agents. Local-first SQLite + vector search. 17 MCP tools.
not reviewed
Growing
B
42
Postgres
Enterprise PostgreSQL MCP server with NL queries, hybrid search (pgvector+BM25), and web UI
not reviewed
Growing
A
43
Citadel
Encrypted-first embedded database with vector search and agent memory, exposed as MCP tools
not reviewed
Growing
B
44
sqlite-graph-memory
Recall over your markdown notes: vector search plus your [[wikilinks]] as a graph, in SQLite.
not reviewed
Growing
A
45
Smart Connections
Local semantic search over Obsidian vaults via Smart Connections embeddings. Multi-vault, private.
not reviewed
Growing
A
46
Pikelet
Search self-contained .pikelet knowledge artifacts over MCP — no vector DB, no embedding API.
not reviewed
Growing
A
47
IT Glue
IT Glue MCP server for MSPs — documents, flexible assets, semantic vector search, RBAC, and BYOK
not reviewed
Growing
A
48
Memory Vault
Local-first AI memory layer with hybrid search. Postgres + pgvector. Self-hosted, MIT.
not reviewed
Growing
B

Or try them without installing any

Add the gateway once and your agent runs search "pinecone", reads the grades, and calls the one it picks. Nothing to install per server, and an F-graded server cannot be reached through it.

Gateway docs

claude mcp add --transport http mcp-market https://gw.mcp.market/mcp

MCP servers for RAG MCP servers for SQL databases MCP servers for PDFs and documents Best Databases MCP servers All topics

MCP servers for vector databases

Before you install one

The vector database servers we would look at first

36 more, graded and ranked

Or try them without installing any

Related