Graphiti MCP Server (Zep)
https://www.getzep.com/product/knowledge-graph-mcp
Tracks how facts change over time; maintains provenance; supports prescribed and learned ontology. Bi-temporal edges (when known vs when valid). Only server that distinguishes "fact learned" from "fact true."
At a glance
- Type
- Temporal KG with bi-temporal edges
- Tier
- T1
- Section
- Claude Code memory mechanisms
- Created
- 2024-08
- Latest release
- v0.29.0 2026-04-27
- License
- Apache-2.0
- GitHub
- 25.7k★ +137/mo Python
- Pricing
- free tier: 2500 messages/mo + 2.5MB graph data; paid: $1.25/1000 messages + $2.50/MB extra; Enterprise custom
- Funding
- not applicable — not commercial
Taxonomy
- storage
- graph
- retrieval
- graph-traversal
- persistence
- long-term
- update
- append-only
- unit
- episode
- governance
- auditable
- conflict
- bi-temporal
When to use
Optimised for: Claude-Code-native + MCP-portable memory
Anti-fit: not for non-Claude-Code clients (some are MCP-portable)
Pros & cons
Pros
MCP exposure of Graphiti's bi-temporal KG gives any agent contradiction-aware memory; first-party vendor support.
Cons
KG memory is heavier than vector for the same data; MCP setup more involved than installing an embedded SDK.
Claims & capabilities
20k★. v1.0 shipped Nov 2025.
Technical surface
- API surface
- MCP (stdio/HTTP)
- Backend storage
- Neo4j
- Deployment
- cloud (Zep managed) or self-hosted via Graphiti OSS + Neo4j/local graph DB
- Embedding model
- multiple supported
- Multi-tenancy
- namespace
- MCP
- native (first-party)
- A2A
- not supported
- OpenTelemetry
- searched not found
Similar systems
Other claude code memory mechanisms in the catalog, ranked by inbound references.
- Official MCP Memory server T1
Knowledge-graph-based persistent memory MCP server — the canonical reference implementation. npx -y @modelcontextprotocol/server-memory ; configurable via MEMORY_FILE_PATH .
- Anthropic Auto Dream T1
Background sub-agent in Claude Code that periodically scans ~/.claude/projects/<project>/memory/ + recent JSONL transcripts via grep, prunes stale notes, deduplicates, resolves contradictions, converts relative dates to absolute timestamps, and rewrites MEMORY.md (kept under 200 lines). Triggers automatically when 24+ hours elapsed AND 5+ sessions completed; manually via /dream.
- Claude Code auto-memory T1
Persistent file-based memory at ~/.claude/projects/<project>/memory/ with a MEMORY.md index plus typed memory files (user / feedback / project / reference). Loaded into context across sessions.
- claude-brain T2
Claude Code plugin packaging of Memvid (see Dedicated memory layers). Wraps the Memvid .mv2 single-file store as a slash-command-installable plugin so per-project memory becomes a git-committable file inside the repo.
- Claude-CursorMemoryMCP T2
FastAPI server. RBAC + API-key auth, encryption, backup/recovery, embedding pipeline with deduplication. Only server in this list with role-based access control over pgvector.
- claude-mem T1
Hooks into Claude Code's session lifecycle; compresses tool outputs into ~500-token observations via Claude's agent SDK. SQLite + FTS5 with optional Chroma vector. "Endless Mode" compresses in-flight to extend session length. Captures what the agent did , not just what the user typed.
Related systems
References (4)
- Anthropic Claude (foundation models) depends on at runtime — adjacent-infrastructure cell: requires Claude Code or MCP client; BYO storage
- Model Context Protocol (MCP spec) depends on at runtime — adjacent-infrastructure cell: requires Claude Code or MCP client; BYO storage
- Neo4j depends on at runtime — backend-storage cell: Neo4j
- Zep & Graphiti builds on — Graphiti is the open-source core; MCP exposure of Graphiti's bi-temporal KG