The memory API your agent stack is missing

Reusable, receipted memory in a few lines.

Turn files, messages, docs, and voice into Unified Memory Objects your agents recall through the REST API + MCP. Capture once, recall many, and show a receipt for every answer.

POST /v1/process
# capture a memory → get a UMO back
curl https://api.memoryintelligence.io/v1/process \
  -H "Authorization: Bearer mi_sk_live_…" \
  -H "Content-Type: application/json" \
  -d '{"content": "Shipping moved to Tuesdays."}'

# → { "umo_id": "8f3f…", "receipt": { "provenance": "verified" } }
# typed SDK coming. REST API works today
import requests

r = requests.post(
    "https://api.memoryintelligence.io/v1/process",
    headers={"Authorization": "Bearer mi_sk_live_…"},
    json={"content": "Shipping moved to Tuesdays."},
)
print(r.json()["umo_id"])   # 8f3f…
// typed SDK coming. native fetch works today
const r = await fetch("https://api.memoryintelligence.io/v1/process", {
  method: "POST",
  headers: {
    "Authorization": "Bearer mi_sk_live_…",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ content: "Shipping moved to Tuesdays." }),
});
const { umo_id } = await r.json();   // 8f3f…
Base URL api.memoryintelligence.ioAuth: Bearer key · free in beta

Start here.

Four doors into the API. Most builders start with the quickstart.

Start with a task.

The core methods, each one call against the REST API.

SDKs & integrations.

Typed SDKs are coming. The REST API and MCP work today.

Python

Coming
# use the REST API today (requests)
Source on GitHub →

JavaScript

Coming
// use the REST API today (fetch)
Source on GitHub →

MCP Server

Live
$ pip install memoryintelligence-mcp
Set up in one command →

Docs your agent can read.

Point Claude, ChatGPT, or Cursor at the whole API and let it write correct integration code on the first try. MemoryIntelligence ships LLM-native docs.

Machine index/llms.txt OpenAPI specmemory-api.yaml