Live on the edge · 5ms cold start

Persistent Memory
for AI Agents

Give your agents long-term memory in two lines of code. Store, semantically retrieve, and forget — backed by vector search, running globally on the edge.

No credit card required · 10,000 memories free every month

How it works

Three verbs. That's the whole API.

Memory is just store, retrieve, and forget. We handle embeddings, vector ranking, caching, and multi-tenant isolation so you don't have to.

💾
POST /memory/store

Store

Send any text plus a user and agent id. We embed it automatically and persist it — metadata included.

🔍
GET /memory/retrieve

Retrieve

Ask a natural-language question. We rank stored memories by semantic similarity and return the most relevant, cached at the edge.

🗑️
DELETE /memory/forget

Forget

Delete a single memory or wipe an entire user/agent scope. Right-to-be-forgotten, one call away.

Integration

So simple it's almost boring.

Drop it into any stack with a single fetch. No SDK required, no infrastructure to run.

  • Automatic embeddings — text in, vector search out. No model to manage.
  • Edge-cached retrieval — sub-100ms responses, globally distributed.
  • Scoped & isolated — every memory is keyed to your API key, user, and agent.
  • Usage built in — track requests, tokens, and latency per key.
agent.js
// 1. Remember something about a user
await fetch("https://agentmemo.dev/memory/store", {
  method: "POST",
  headers: {
    Authorization: "Bearer am_sk_your_key",
    "Content-Type": "application/json"
  },
  body: JSON.stringify({
    user_id: "user_123",
    agent_id: "support_bot",
    content: "Prefers email. On the Pro plan."
  })
});

// 2. Recall it later, semantically
const res = await fetch(
  "https://agentmemo.dev/memory/retrieve?" +
  "user_id=user_123&q=how to contact them",
  { headers: { Authorization: "Bearer am_sk_your_key" } }
);
// → "Prefers email. On the Pro plan."  (score 0.62)
Pricing

Start free. Scale when you're ready.

No setup fees. No per-seat pricing. Pay for memory, not meetings.

Free
$0 / month

For prototypes and side projects.

  • 10,000 memories / month
  • Semantic retrieval included
  • Edge caching
  • 1 API key
  • Community support
Get started free
Most popular
Pro
$19 / month

For production agents at scale.

  • Unlimited memories
  • Priority embedding queue
  • Unlimited API keys & scopes
  • Usage analytics dashboard
  • Email support
Upgrade to Pro

Ship agents that remember.

Grab an API key and store your first memory in under a minute.