From Forgetful to Brilliant.
Watch how AutoMem transforms your AI from a blank slate into a context-aware assistant.
Universal I/O
Compatible with Claude Code, Cursor, ChatGPT, OpenClaw, Antigravity, and any MCP client. Local or remote — works everywhere via MCP and Streamable HTTP.
Graph Engine
FalkorDB + Qdrant hybrid core. Maps relationships between memories, not just vector similarity.
Instant Recall
Sub-50ms latency. Your agent knows what you know before you finish typing the prompt.
Where AutoMem Works
Wherever you work — Claude, Cursor, Copilot, Codex, ChatGPT, Windsurf, or another MCP-compatible client — AutoMem can sit behind it. Local tools use stdio MCP; cloud tools can connect through Streamable HTTP.
Claude Desktop
stdio MCP
Claude Code
stdio MCP
Cursor
stdio MCP
GitHub Copilot
stdio MCP via VS Code
Codex
stdio MCP
Windsurf
stdio MCP
ChatGPT
Streamable HTTP / remote MCP
Any MCP-compatible client
MCP or Streamable HTTP
AutoMem supports both MCP and Streamable HTTP, so your memory store can follow you between local coding assistants and browser-hosted AI tools without becoming tied to one vendor.
Dreams while you dream.
AutoMem runs quietly in the background, stitching every thought into a retrieval graph. It notices patterns, links context, and keeps the right edges warm—without you ever thinking about it.
0.15.2 • Recall Quality & Tag Ops
Keyword scoring for vector results with a softer adaptive floor. Expansion candidates now bypass the tag filter so graph walks aren't cut off at the seed set. Memory-by-tag pagination and bulk delete endpoints. LoCoMo judge runs hardened against flaky rate limits.
MCP 0.14.0 • OpenClaw Plugin-First & Antigravity
Plugin-first memory parity for OpenClaw with shared policy across plugin, MCP, and skill modes. Google Antigravity setup templates shipped. Claude Code plugin deprecated in favor of the CLI installer. Windows Python wrapper for hook scripts.
0.15.0 • Relationship Engine & Benchmarks
Optimized relationship taxonomy for cleaner graph traversal. LoCoMo cat5 multi-hop judge for automated recall quality scoring. priority_ids now boosts relevance instead of hard-filtering. Embedding dimension validation catches provider mismatches at startup.
MCP 0.13.0 • OpenClaw Overhaul & Cursor Rules
Complete OpenClaw plugin rewrite — native MCP integration and new skill setup. Split global vs. project rules for Cursor. Hook stdin parsing with truncation guards and dedup checks.
0.14.0 • Docker & Hardened Deployments
Official Docker build workflow. QDRANT_HOST + QDRANT_PORT for flexible Qdrant config. Stateless MCP bridge transport for resilient connections. Consolidation overhaul: reduced decay rate, importance floor, archived memory filtering. Voyage AI set as recommended embedding default.
0.13.0 • Adaptive Recall & Graph Viewer
min_score threshold with adaptive floor filtering — low-relevance results get cut automatically. LoCoMo benchmark harness for rapid experiment iteration. Graph viewer externalized as standalone module.
0.12.0 • Recall Quality Lab & JIT Enrichment
Just-in-time enrichment on recall for higher-quality results. Recall Quality Lab for data-driven recall optimization. LongMemEval benchmark harness (ICLR 2025) for measuring memory quality. relevance_score now synced to Qdrant payload for accurate vector search.
0.11.0 • Embedding Providers & Memory Lookup
Voyage AI embedding provider for high-accuracy recall. OpenAI-compatible providers via OPENAI_BASE_URL (LM Studio, Ollama, xAI, etc.). GET /memory/<id> endpoint for direct memory lookup. Hardened time parser for numeric timestamps.
0.10.1 • Stability Fixes
Temporal query bound to 7-day window with timeout to prevent runaway queries. Skip temperature param for o-series and gpt-5 models.
0.10.0 • Visualization & Observability
Interactive graph visualization API. automem-watch real-time observability with SSE streaming and TUI. Ollama embedding provider for fully local deployments. Memory content size governance with auto-summarization. Streamable HTTP transport for MCP connections.
MCP 0.12.0 • OpenClaw Integration
Native OpenClaw integration replacing mcporter dependency chain. Direct HTTP API calls to AutoMem with new SKILL.md template. Simplified setup and troubleshooting documentation.
MCP 0.11.0 • Claude Code Plugin
Claude Code plugin with automatic memory capture hooks. Session start recall, session end summary. Cursor hooks integration for automatic capture. Integration tests and CLI smoke tests.
Deploy Once. Remember Everywhere.
Your memory infrastructure should be as portable as your code. Start with a hosted pod, launch on Railway, or run the full stack yourself.
InstaPods
Deploy AutoMem in about 30 seconds on the Grow plan with HTTPS, custom domains, SSH access, and generated MCP config.
> $15/mo flat
Railway
One-click production deployment with usage-based pricing, generated environment variables, and a public HTTPS domain.
> Ready in ~60 seconds
Self-hosted
Run the Docker stack locally or on your own VPS. Bring your own domain, HTTPS, operations, and SSH policy.
npm install
Provider Comparison
Choose the deployment path by how much platform work you want AutoMem to do for you.
| Capability | InstaPods | Railway | Self-hosted / Docker |
|---|---|---|---|
| Deploy time | 30 seconds | 60 seconds | Manual |
| Pricing | $15/mo flat | Usage-based | VPS or local cost |
| Config | Generated MCP config | Generated env vars | Manual config |
| HTTPS | Included | Included | Manual HTTPS |
| SSH | SSH access | Not the focus | Host-dependent SSH |