Free to start — from $9.9/mo
Your AI agent
remembers everything
Cloud memory for Claude Code, Cursor, and any MCP-compatible agent. Context persists across sessions, projects, and teams.
# Agent recalls context from last session
memory_recall
> Found 3 memories for my-project:
decision — Use PostgreSQL, not MongoDB
architecture — REST API with JWT auth
feedback — User prefers minimal PRs
# Works across sessions automatically
memory_store
> Stored: "Deploy uses Docker + Nginx"
What you get
Persistent memory
Decisions, feedback, architecture choices — stored once, recalled everywhere. No more repeating context.
Multi-project isolation
Each repo gets its own memory scope. Switch projects without cross-contamination.
Topic + importance
Memories are organized by topic and ranked by importance. Retrieval is weighted, not random.
MCP native
14 tools via Model Context Protocol. Works with Claude Code, Cursor, Windsurf, and any MCP client.
Tenant isolation
Your data is isolated per team. JWT auth, encrypted at rest, no cross-tenant access.
Semantic search
BGE-M3 embeddings for cross-language retrieval. Find memories by meaning, not just keywords.
Two files. Done.
Drop two config files into your project. No SDK, no wrapper code, no boilerplate.
Get your API token
Sign up free, copy the token from dashboard.
Add the config
.mcp.json + .claude/settings.local.json
Start coding
Your agent now has persistent memory. It just works.
Let your agent do it
Send the setup guide to your agent and it will configure everything automatically:
https://agent-memo.ai/setup.mdViewREST API
https://api.agent-memo.ai| Method | Endpoint | Description |
|---|---|---|
| GET | /api/memories | List all memories |
| POST | /api/memories | Create new memory |
| GET | /api/memories/:id | Get memory by ID |
| PUT | /api/memories/:id | Update memory |
| DELETE | /api/memories/:id | Delete memory |
| POST | /api/memories/search | Semantic search with filters |
| GET | /api/memories/topics | List topics with counts |
| POST | /api/kg/add | Add knowledge graph fact |
| POST | /api/kg/query | Query entity relationships |
Authorization: Bearer YOUR_API_KEYFrequently asked questions
What is Agent-Memo.AI?
A cloud memory service for AI coding agents. It gives agents like Claude Code, Cursor, and Windsurf persistent memory across sessions — decisions, preferences, and project context are stored and recalled automatically.
How does it work with Claude Code?
Via the Model Context Protocol (MCP). Add a .mcp.json config file to your project, and Claude Code gets 14 memory tools — store, recall, search, knowledge graph, and more. Hooks auto-save context.
Is it free?
Yes — the Free plan includes 3 projects, 1,000 memories, and 1,000 API calls/month. No credit card required. Pro starts at $9.9/mo for 20 projects and 50K memories. See our pricing page for details.
Which AI agents are supported?
Any MCP-compatible agent: Claude Code, Cursor, Windsurf, and custom MCP clients. If your agent supports MCP, it works with Agent-Memo.
Is my data secure?
All data is isolated per team with JWT auth. Encrypted in transit (SSL/TLS), passwords hashed with bcrypt, no cross-tenant access. We don't sell or share your data.
Do I need to install an SDK?
No. The MCP client runs via npx — no global installation, no SDK, no boilerplate. Just two config files in your project root.
Simple, transparent pricing
Start free, upgrade when you need more.
Free
Get started for free
- 3 projects
- 1,000 memories
- 1,000 API calls / month
- 100 KG facts
- Semantic search
- 14 MCP tools
Pro
$99/yr (17% off)
- 20 projects
- 50,000 memories
- 50,000 API calls / month
- 5,000 KG facts
- Per-project instructions
- Email support
Team
$499/yr (17% off) + $7/seat/mo
- Unlimited projects & memories
- Unlimited API calls & KG facts
- $7/seat/mo (up to 50 seats)
- Shared memories & permissions
- Per-project instructions
- Priority support