Memobase
https://github.com/memodb-io/memobase
User-profile-based long-term memory. Per-user profile + event timeline. API + Python/Node/Go SDKs + MCP server. Profile-centric rather than fact-centric.
At a glance
- Type
- User profile + event timeline
- Tier
- T2
- Section
- Dedicated memory layers
- Created
- 2024-09
- Latest release
- no releases (no tagged releases on GitHub as of research date)
- License
- Apache-2.0
- GitHub
- 2.7k★ +10/mo Python
- Pricing
- Free + paid (Memobase Cloud free tier for testing; pay-as-you-go for production)
- Funding
- not applicable — not commercial
Taxonomy
- storage
- relational
- retrieval
- injection
- persistence
- long-term
- update
- extraction
- unit
- profile
- governance
- inspectable
- conflict
- none
When to use
Optimised for: user-profile + event-timeline (companion-AI persona)
Anti-fit: not for non-Claude-Code clients (some are MCP-portable)
Pros & cons
Pros
User-profile-centric model is closer to how product teams think about user memory than fact-extraction approaches; cross-tool memory passport via MCP.
Cons
Smaller community than Mem0/Letta; profile-centric framing doesn't fit purely conversational use cases as well.
Claims & capabilities
Sub-100ms retrieval. Cross-tool memory passport for Claude / Cursor / MCP-compatible agents.
Technical surface
- API surface
- REST, SDK: Python, JS/TS
- Backend storage
- Postgres + pgvector
- Deployment
- Both (self-hosted Docker + Memobase Cloud managed)
- Embedding model
- multiple supported
- Multi-tenancy
- namespace (user_id)
- MCP
- native (first-party) — memobase MCP
- A2A
- not supported
- OpenTelemetry
- no
Similar systems
Other dedicated memory layers in the catalog, ranked by inbound references.
- Mem0 T1
Universal memory layer for AI agents. Three concurrent stores (vector + graph + KV); LLM-extracted facts; concurrent retrieval via ThreadPoolExecutor.
- Zep & Graphiti T1
Bi-temporal knowledge graph (event time + ingestion time). Strong on chronological reasoning and contradiction tracking. Graphiti is the open-source core.
- Cognee T1
"Extract–Cognify–Load" pipeline that turns raw input into a typed, queryable knowledge graph for agent recall.
- Hindsight (Vectorize) T1
Standalone memory service from Vectorize. Open source. Biomimetic four-network design (World, Bank, Observation, Opinion). Ships an MCP memory server.
- Memvid T2
Single-file memory layer (one .mv2 file). No DB, no server. Append-only sequence of immutable Smart Frames with timestamps + checksums. Native Rust core (rewritten from Python).
- Supermemory T1
Memory engine with API, app, browser extension, and MCP server. Extracts facts, tracks updates, resolves contradictions, auto-forgets expired info. Plugins for Claude Code, OpenCode, OpenClaw, Hermes.
Related systems
References (2)
- Model Context Protocol (MCP spec) depends on at runtime — r Claude / Cursor / MCP-compatible agents.
- pgvector depends on at runtime — backend-storage cell: Postgres + pgvector