About this tool

The IteraTools Vector Memory Search API retrieves documents from your namespace by semantic similarity. It embeds your query with OpenAI text-embedding-3-small and ranks stored documents by cosine similarity. Returns the top_k most relevant results with their scores (0–1, higher = more similar), original text, and metadata. Namespace isolation ensures your data is private to your API key.

Quick Start

curl -X POST https://api.iteratools.com/memory/search \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "namespace": "my-agent",
    "query": "What does the user prefer?",
    "top_k": 3
  }'

Response

{
  "ok": true,
  "data": {
    "results": [
      {
        "id": "fact-1",
        "text": "The user prefers dark mode and uses Python.",
        "score": 0.92,
        "metadata": {"source": "preferences"}
      },
      {
        "id": "fact-3",
        "text": "User is building a rhythm game called Sambamancer.",
        "score": 0.71,
        "metadata": {}
      }
    ],
    "count": 2
  }
}

Python Example — RAG Pipeline

import requests

def memory_search(namespace: str, query: str, api_key: str, top_k: int = 5) -> list:
    res = requests.post(
        "https://api.iteratools.com/memory/search",
        headers={"Authorization": f"Bearer {api_key}"},
        json={"namespace": namespace, "query": query, "top_k": top_k}
    )
    return res.json()["data"]["results"]

# RAG: retrieve context before answering
query = "What programming language does the user prefer?"
results = memory_search("user-profile", query, "YOUR_KEY", top_k=3)

context = "\n".join([f"- {r['text']}" for r in results if r['score'] > 0.7])
print(f"Relevant context (score ≥ 0.7):\n{context}")

# Feed to LLM
prompt = f"Based on user context:\n{context}\n\nAnswer: {query}"

Request Body

namespace	string (required)	Namespace to search in
query	string (required)	Natural language search query
top_k	integer (optional)	Max results to return (default: 5, max: 100)

Details

Endpoint	POST /memory/search
Price	$0.002 / request
Similarity metric	Cosine similarity (0–1)
Embedding model	OpenAI text-embedding-3-small (1536 dims)
Isolation	Per API key (namespaces prefixed automatically)
Auth	Bearer token or x402 micropayment
Base URL	https://api.iteratools.com

Use Cases

🤖 RAG pipelines — retrieve context before passing to an LLM
💬 Conversational memory — find relevant past interactions
📖 Knowledge base QA — search indexed documents by meaning
🎯 Recommendation — find similar items, products, or content
🔗 Agent tool use — give LLMs access to semantic memory

Full Documentation Memory Upsert → Memory Clear → Browse All Tools