Spomory

Spomory is a personal AI memory engine that builds an explainable knowledge graph to share context across AI tools.

Spomory is a personal AI memory engine that lets your context follow you across different AI applications. It extracts text into structured subject, predicate, and object triples, merging them incrementally into a dual-layer memory graph rather than a raw text blob. When you ask a question, it matches your query against the triples and runs personalized PageRank over the graph to retrieve connected multi-hop context. A reinforcement-learned policy decides whether to keep, update, or retire memories. It runs locally via an MCP server with a SQLite database, or via a cloud-hosted endpoint backed by Postgres. It is designed for developers and technical users who want memory sovereignty and verifiable retrieval.

Spomory at a glance

Best for: Software engineer
Runs on: Web, macOS, Windows, Linux, iOS, Android, Chrome extension, API, VS Code, JetBrains
Pricing: Free during early access; local mode permanently free
Released: October 2026
Category: AI coding tool
Website: spomory.yliuai.com

Key features

Extracts text into structured subject-predicate-object triples
Merges new facts incrementally into a dual-layer memory graph
Performs personalized PageRank retrieval for multi-hop context
Uses a reinforcement-learned policy to manage memory updates
Runs locally via an MCP server with SQLite storage
Provides a cloud-hosted endpoint backed by Postgres
Exports full memory graphs in an open JSON-LD-style format
Performs permanent deletions with an audit trail

Use it for

Sharing project context across different AI applications
Storing engineering details locally on a personal disk
Retrieving multi-hop facts through graph traversal
Managing and retiring outdated memory entries automatically

Alternatives to Spomory

claude-mem — It shares persistent, searchable agent memory across sessions and devices.
DocuMemory — Turns messy AI chat transcripts into clean, reusable context for future sessions.
Engram — Single Go binary with zero dependencies and built-in full-text search.
AI Memory — Shared memory for coding agents that works across machines and team members.
M-Flow — It achieves a 93.7 percent top-one hit rate using a conical memory architecture.
Memorax Code — Carry project architecture and verified engineering decisions across coding sessions.
Pieces — Recovers past decisions and supplies historical context to AI tools.
IAI Memory Engine — Provides persistent, local, and encrypted memory for AI agents without APIs.
Cognee — Persistent memory for AI agents using graph, vector, and relational retrieval.
memU — It gives every connected agent a shared and inspectable memory layer.
ContextMCP — Maintains fresh, accurate context for AI agents using AST-aware chunking.
MemOS — It delivers ultra-persistent memory, hybrid-retrieval, and token savings.

All Spomory alternatives