Integrations

xMem is designed to be modular and pluggable. You can integrate with a wide range of LLMs, vector stores, and session stores. Here are some common integrations:

LLM Integrations

• Llama.cpp
• Ollama
• Mistral
• HuggingFace Inference API
• OpenAI
• Gemini (Google AI)

import { LlamaCppAdapter } from class="text-amber-300">'./adapters/llamaCpp';
orchestrator.registerProvider(class="text-amber-300">'llm', class="text-amber-300">'llama', new LlamaCppAdapter({ apiUrl: class="text-amber-300">'http:class="text-slate-400">//localhost:8080' }));

Vector Store Integrations

• ChromaDB
• Qdrant
• Pinecone
• MockVectorStore (for dev/testing)

import { ChromaDBAdapter } from class="text-amber-300">'./adapters/chromadb';
orchestrator.registerProvider(class="text-amber-300">'vector', class="text-amber-300">'chromadb', new ChromaDBAdapter({ url: class="text-amber-300">'http:class="text-slate-400">//localhost:8000', collection: class="text-amber-300">'my_collection' }));

Session Store Integrations

• PostgreSQL
• MongoDB

Tip: You can register multiple providers of each type and select which to use per operation.

n8n Integration

n8n is an open-source workflow automation tool. You can connect n8n to xMem using the built-in HTTP Request node to automate memory upserts, queries, and more. No custom plugin is required—just use your xMem API endpoints!

1. In n8n, add an HTTP Request node.
2. Set the Method to POST (or GET as needed).
3. Set the URL to your xMem API endpoint (e.g., https://your-xmem.com/api/session-memory).
4. Configure authentication (API key or session, as required).
5. Set the Body to JSON and provide the required fields (see below).

Example: Upsert Session Memory

POST https://your-xmem.com/api/session-memory
Content-Type: application/json
Authorization: Bearer YOUR_API_KEY

{
  "sessionId": "session-123",
  "content": "Remember this fact.",
  "role": "user"
}

n8n HTTP Request Node Config

{
  "url": "https://your-xmem.com/api/session-memory",
  "method": "POST",
  "headers": {
    "Content-Type": "application/json",
    "Authorization": "Bearer YOUR_API_KEY"
  },
  "body": {
    "sessionId": "session-123",
    "content": "Remember this fact.",
    "role": "user"
  },
  "bodyContentType": "json"
}

Tip: You can use any xMem API endpoint in n8n. For advanced workflows, chain multiple HTTP Request nodes or use n8n's built-in logic to automate your LLM memory orchestration!