GraphThink

Sovereign graph memory for AI agents. Self-hosted. Zero API costs.

Your agent loses context between sessions. Existing memory APIs charge per token and lock you into their cloud. GraphThink gives you persistent, semantic memory on your own infrastructure.

Quick Start

pip install graphthink

from graphthink import GraphThink

gt = GraphThink()

# Store memories
gt.store("user", "I prefer FastAPI over Django")
gt.store("user", "My project is a customer support bot")

# Search semantically — works across sessions
results = gt.search("what framework do I use")
# → [{"content": "I prefer FastAPI over Django", "score": 0.89}]

# See what's in your graph
print(gt.stats())
# → {"conversation": 2, "message": 3, "entity": 4, ...}

Why GraphThink?

Feature	Mem0	Zep	GraphThink
Graph relationships	❌	❌	✅
Self-hosted	❌	❌	✅
Zero per-query costs	❌	❌	✅
Free web search	❌	❌	✅
API key required	✅	✅	❌ (open)

Architecture

Your App ──→ GraphThink Client ──→ GraphThink API ──→ Memgraph
(pip install)   (localhost:18788)     (your Docker)

All data stays on your machine. No third-party servers.

Use Cases

• Customer support bots — remember user history across tickets
• Personal AI assistants — persist preferences indefinitely
• Multi-agent systems — share context between agents
• Cost-sensitive teams — eliminate per-token memory costs

Self-Hosted Setup

# Start Memgraph (you need this once)
docker run -d --name memgraph -p 7687:7687 memgraph/memgraph-platform

# Start GraphThink API
graphthink serve

# That's it. Your memories, your infrastructure.

License

MIT — free for personal and commercial use.