The Memory API

Give any application persistent AI memory with a simple REST API. Ingest, search, retrieve, and consolidate — all over HTTPS.

Get API Key Read Docs
API Operational
AES-256-GCM Encrypted
4 LLM Providers
Base: https://igenius-memory.online/v1

Get Your API Key

Create an account to receive your API key instantly.

Already have an account?

Endpoints
Everything You Need

Base URL: https://igenius-memory.online/v1

POST
/ingest
Process a user or agent message into persistent memory via AI classification.
GET
/briefing
Get the consolidated intelligence briefing for your agent's session start.
POST
/consolidate
Merge accumulated persistent memories into a master briefing report.
POST
/process
Run trigger word detection on text, catching passwords, URLs, and configs.
GET
/memories/search?q=
Semantic search across all memory layers. Returns ranked results.
POST
/store
Directly save a memory to a specific layer with custom title and tags.
GET
/memories/review
Triage short-term catches — review before promoting or discarding.
POST
/memories/{id}/promote
Promote a short-term memory to long-term permanent storage.
DELETE
/memories/{id}
Permanently remove a memory by ID.
POST
/pin
Pin a fact permanently. Never expires, always encrypted. Requires user confirmation.
Quick Start
Start in 30 Seconds
python — ingest + briefing
import httpx BASE = "https://igenius-memory.online/v1" HEADERS = {"X-API-Key": "ig_your_key_here"} # Ingest a user message resp = httpx.post(f"{BASE}/ingest", json={ "message": "We decided to use PostgreSQL instead of SQLite", "role": "user" }, headers=HEADERS) print(resp.json()) # {"ingested": true, "memory_id": 42, "title": "Database Decision", ...} # Get briefing at next session start briefing = httpx.get(f"{BASE}/briefing", headers=HEADERS) print(briefing.json()["briefing"]) # "### Recent Decisions\n- Switched from SQLite to PostgreSQL..."
New Feature
Visual Tools

Render, screenshot, and analyze any URL with a local vision model. Available as MCP tools or via the VS Code extension.

📊

visual_report

MCP Tool

Render a URL → take a screenshot → analyse with the vision model → return a full UI/UX report with professional score, top-3 fixes, and CSS/HTML suggestions.

Parameters:
  url (string, required) — URL to render
  viewport_width (int, optional) — default 1280
  viewport_height (int, optional) — default 800
📸

visual_screenshot

MCP Tool

Render a URL and return the screenshot as a base64-encoded PNG. No vision analysis — useful for embedding or passing to your own model.

Parameters:
  url (string, required) — URL to render
  viewport_width (int, optional) — default 1280
  viewport_height (int, optional) — default 800

🛠️ Requires: pip install "igenius-mcp[visual]" + playwright install chromium + a vision model (local via LM Studio, or cloud via OpenAI/Anthropic/Google).

Pricing
Simple, Transparent Pricing

Start free. Scale when you need to.

Free

$0/mo
  • ✓ 1,000 requests/week
  • ✓ 4 memory layers
  • ✓ 17 MCP tools
  • ✓ 30-day history
  • ✓ Community support
Get Free Key →

Pro

$19/mo
  • ✓ 50,000 requests/day
  • ✓ 4 memory layers
  • ✓ 17 MCP tools
  • ✓ 365-day history
  • ✓ Priority support
  • ✓ Webhook notifications
Upgrade to Pro →

Enterprise

Custom
  • ✓ Unlimited requests
  • ✓ Custom memory layers
  • ✓ Dedicated infrastructure
  • ✓ Unlimited history
  • ✓ Dedicated support
  • ✓ SLA guarantee
Coming Soon