whereismyai
Graphmemory
Graphmemory is a bitemporal memory service for AI agents that provides long-term memory with six-variable decay and time-graded recall.
Graphmemory is a bitemporal memory service designed for AI agents to accumulate, forget, and evolve memories over time. It uses a temporal knowledge graph on top of Neo4j, PostgreSQL with pgvector, and Redis. The system supports a six-variable importance and decay model, time-graded recall, TTL-based expiry, deduplication, and competitive retrieval. It includes an append-only episode ledger, a built-in Model Context Protocol server, and a minimal web graphical user interface with a REST application programming interface. The software targets developers building AI agents who need persistent, time-aware memory storage using Neo4j, PostgreSQL, and Redis databases.
Graphmemory at a glance
- Best for: Software engineer
- Runs on: API, Web, self-hosted
- Pricing: Free
- Released: October 2026
- Category: AI agent tool
- Website: github.com/gh9ko8y/graphmemory
Key features
- Bitemporal facts with valid and transaction time axes
- Six-variable importance and decay model
- Time-graded recall across multiple intervals
- Time-to-live based expiry and expiration handling
- Competitive retrieval layer with relative thresholds
- Model Context Protocol server for agent integration
- Append-only episode ledger for tracking original input
- Minimal web graphical user interface and REST API
Use it for
- Storing long-term memory for AI agents
- Tracking facts with valid and transaction time axes
- Managing memory expiry and time-graded recall
- Running competitive retrieval to surface relevant facts
Alternatives to Graphmemory
- Supermemory — Integrates long-term memory into AI coding agents with a single prompt.
- AgentMemory — It runs locally with zero external databases and achieves high retrieval recall.
- ReMe — It enables local-first, file-native long-term memory sharing across multiple AI agents.
- Memori — Reduces LLM token usage and spend by over 97% with targeted recall.
- MemPalace — Achieves high recall with local-first storage and zero API calls.
- MemMachine — Provides long-term persistent memory for LLM-based agents.
- Mem0 — It provides drop-in persistent memory for AI agents without changing pipelines.
- memU — It gives every connected agent a shared and inspectable memory layer.
- M-Flow — It achieves a 93.7 percent top-one hit rate using a conical memory architecture.
- Agentmemoryrepo — Enables agents to navigate memories like a wiki and share knowledge across sessions.
- LongMemory — Provides immutable, temporal, and governed recall instead of simple vector retrieval.
- MemOS — It delivers ultra-persistent memory, hybrid-retrieval, and token savings.
All Graphmemory alternatives