Memro
Memro is a long-term memory layer for AI agents. It stores interactions as typed memories, consolidates them into lasting facts, follows relationships in a reasoning graph, detects stale or conflicting facts, and returns relevant context before an agent prompts a model.
At a Glance
- Individual developers and hobbyists
- Engineering teams building multi-agent systems and services
- Businesses and enterprise agent workloads
- Teams using AI coding assistants
- +2 more
AI Tools by Memro
(1)Memro
Long Term Memory for AI Agents
Discussions
No discussions yet
Be the first to start a discussion about Memro
Latest News
Memory corrections and history, point-in-time recall, stale-fact ranking, improved consolidation, Claude Code plugin, and memro-cli doctor
Organization accounts, tenant keys, member usage, improved keyword search, and upgraded WebGL memory graph
Faster cached dashboard memory graph and improved Memro Cloud sign-in presentation
Temporal fact updates, multi-session question aggregation, and memory graph in the dashboard and API
Products & Services
Managed long-term memory service for AI agents, with hosted API/MCP access, managed PostgreSQL and pgvector, daily backups, dashboards, agent identities, connectors, and commercial plans.
Source-available Memro server and PostgreSQL/pgvector deployment runnable with Docker Compose; free for non-commercial use under PolyForm Noncommercial 1.0.0.
Model Context Protocol server that exposes persistent memory to Claude Code, Cursor, Codex, Claude Desktop, and other MCP clients.
Command-line connector, including npx memro-cli connect and status/doctor/disconnect workflows, that configures supported AI coding tools.
Market Position
Memro positions itself as source-available agent-memory infrastructure rather than another vector store: it combines typed memories, consolidation, a multi-hop reasoning graph, contradiction and temporal handling, agent-owned identity, and cross-tool MCP access. Its own comparison material names Mem0, Zep, Supermemory, Letta, Graphiti, and conventional vector databases as alternatives or adjacent approaches.
Business Model
Revenue Model
Managed Memro Cloud subscriptions for commercial use, priced by stored tokens, monthly queries, and monthly ingest. Self-hosting is free for non-commercial development under PolyForm Noncommercial; commercial products use Memro Cloud or a commercial license. Memro also documents Dodo Payments for human subscriptions and x402/Base USDC payments for autonomous agents.
Pricing Tiers
100,000 stored tokens, 5,000 queries/month, 100 MB monthly ingest; CLI and MCP support; PolyForm Noncommercial self-hosting.
1,500,000 stored tokens, 25,000 queries/month, 1 GB monthly ingest; all platform features.
6,000,000 stored tokens, 150,000 queries/month, 5 GB monthly ingest; all platform features.
25,000,000 stored tokens, 1,000,000 queries/month, 20 GB monthly ingest; all platform features.
Target Markets
- Individual developers and hobbyists
- Engineering teams building multi-agent systems and services
- Businesses and enterprise agent workloads
- Teams using AI coding assistants
- Developers and organizations needing self-hosted, data-owning memory infrastructure
- Commercial AI product builders
- AI agents and autonomous copilots
- AI coding assistants sharing memory across Claude Code, Cursor, Codex, VS Code, Gemini CLI, Antigravity, OpenCode, Windsurf, and Claude Desktop
- Customer-support copilots
- Research assistants that retain linked findings across sessions
- Sales copilots that remember stakeholders, promises, and account changes
- Tutoring and learning assistants