Docs · Quickstart

Two lines to memory.

Get a key, point your client

Create an account, copy your ick_ key, and set your client's base URL to the gateway. Memory, security, and caching switch on for every call. No SDK, no rewrite.

# 1 — teach it something (any session)
system: "Our launch window is Q3. Refunds within 30 days."

# 2 — a later session recalls it automatically
user: "When do we launch?"
← "Q3." (you never re-sent it)

Anthropic-style

ANTHROPIC_BASE_URL=https://www.iconia.tech/v1
ANTHROPIC_API_KEY=ick_your_key

Works in any client that reads a custom Anthropic base URL. Some IDE agents pin their own endpoint and won’t honor ANTHROPIC_BASE_URL — if yours does, reach out and we’ll help you wire it up.

OpenAI-compatible

from openai import OpenAI

client = OpenAI(
  base_url="https://www.iconia.tech/v1",
  api_key="ick_your_key"
)

BYOK — bring your own key

Prefer to keep billing on your own provider account? Register your provider key and it becomes your Iconia identity. Iconia injects memory and screens the request, then forwards through your credentials. Your key stays yours.

Response headers

Every response carries receipts — what happened, and how fast.

x-iconia-augmented1 if memory was injected, else 0
x-iconia-items-injectedhow many memories were added to the call
x-iconia-retrieval-msretrieval time in milliseconds
x-iconia-cachehit-semantic · miss · off
x-iconia-cache-latency-mstotal time when served from cache

Semantic cache

Repeated meaning shouldn't cost full price twice. Opt in per request with x-iconia-cache: semantic — same-meaning questions answer from cache without a model call. Opposite meanings and different questions always miss and go to the model. When your memories change, affected answers invalidate automatically.