Mengram vs claude-mem
claude-mem is a simple Claude Code hook that saves conversations to Markdown files. Mengram provides a complete memory system: auto-save, auto-recall on every prompt, cognitive profile on session start, and 3 memory types.
What claude-mem does well
- Simple and easy to understand
- Works as a Claude Code hook
- Saves conversations to local Markdown files
- No API key or account needed
What claude-mem doesn't have
- No auto-recall — saves but never retrieves
- No cognitive profile loading on session start
- No semantic search across memories
- No episodic or procedural memory
- No cross-tool support (only Claude Code)
- No MCP server, no LangChain/CrewAI integration
- No knowledge graph or entity extraction
- Local-only — no cloud sync or multi-device
Feature comparison
| Mengram | claude-mem | |
|---|---|---|
| Semantic memory (facts) | ✅ | ❌ |
| Episodic memory (events) | ✅ | ❌ |
| Procedural memory (workflows) | ✅ | ❌ |
| Self-improving procedures | ✅ | ❌ |
| Cognitive Profile | ✅ | ❌ |
| Multi-user isolation | ✅ | ❌ |
| Knowledge graph | ✅ | ❌ |
| MCP server | ✅ | ❌ |
| Import (ChatGPT, Obsidian) | ✅ | ❌ |
| Self-hostable | ✅ | ✅ |
| Price | Free | Free (local) |
When to choose what
Choose claude-mem if...
Simple conversation logging to local files. Good if you just want a text record of past Claude Code sessions.
Choose Mengram if...
Full memory loop — auto-save, auto-recall, cognitive profile. Claude knows who you are, what you worked on, and brings relevant context on every prompt. Works across Claude Code, MCP, LangChain, CrewAI, and REST API.