Memoria (MatrixOrigin)
Memory layer from MatrixOrigin. Strong-retrieval / clear-reader-separation design.
Active · Verified 2026-10-22 · fresh as of 2026-10-22 · 35 cited facts, 1 with a third-party source · 1 recorded change · Suggest a correction
At a glance
- Type
- Retrieval-strong memory layer vendor fix
- Tier
- T2
- Section
- Dedicated memory layers
- Created
- searched not found fix
- Latest release
- v0.5.2 (Sep 11; page shows no year, presumably 2026) repo fix
- License
- Apache-2.0 repo fix
- GitHub
- matrixorigin/memoria — ~607 stars repo fix
- Pricing
- searched not found fix
- Funding
- searched not found fix
- Freshness
- 2026-10-22 fix
- Layer
- agent-infra repo fix
- Licensing
- open-source repo fix
- Protocols
- mcp-server repo fix
- Developer activity
- OpenRank 2.69 (12-mo avg 5.2) — 2026-05 third-party fix
Memory taxonomy
- storage
- hybrid
- retrieval
- similarity
- persistence
- long-term
- update
- extraction
- unit
- fact
- governance
- inspectable
- conflict
- none
When to use
Optimised for: retrieval quality ("retrieval-strong")
Anti-fit: searched not found
Pros & cons
Pros
Backed by MatrixOrigin's HTAP DB engine — memory benefits from mature transactional substrate.
Cons
Tied to MatrixOrigin database; primarily Chinese-market positioning.
Claims & capabilities
Reports 88.78% overall on LongMemEval-S; near-perfect single-session recall; strong on multi-session reasoning, temporal reasoning, and knowledge updates.
Technical surface
- API surface
- searched not found
- Backend storage
- searched not found
- Deployment
- Both — Memoria Cloud (thememoria.ai, remote mode) or self-hosted Docker Compose with MatrixOne (embedded mode); Kubernetes documented
- Embedding model
- searched not found
- Multi-tenancy
- searched not found
- MCP
- yes — native MCP server (memoria mcp) for any MCP-compatible agent
- A2A
- not supported
- OpenTelemetry
- not documented publicly
Alternatives to Memoria (MatrixOrigin) → live options ranked by shared design, licence, hosting and MCP support
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.
- Letta / MemGPT T1
Originated as MemGPT (UC Berkeley), where the model pages its own memory between context, recall and archival tiers. The product now centres on Letta Code, a stateful agent harness whose agents rewrite their own memory, skills and prompts over time.
- Memori T2
Structured, persistent memory layer for AI agents. Open source.
Sources (34 cited facts, 1 third-party)
Each fact links to the page it came from; the badge says who published it (how to read sources).
- What it is Retrieval-strong memory layer vendor
- What it does & what's distinct Memory layer from MatrixOrigin. Strong-retrieval / clear-reader-separation design. vendor
- Claims / benefits Reports 88.78% overall on LongMemEval-S; near-perfect single-session recall; strong on m… vendor
- Modalities text vendor
- Latest release v0.5.2 (Sep 11; page shows no year, presumably 2026) repo
- License Apache-2.0 repo
- GitHub matrixorigin/memoria — ~607 stars repo
- Customers / scale community ppl Database developers, open-source repo
- Data handling Data stored in MatrixOne with MVCC/CoW architecture; on-prem option available repo
- Deployment Both — Memoria Cloud (thememoria.ai, remote mode) or self-hosted Docker Compose with Mat… repo
- HQ China (MatrixOrigin) repo
- Founders / pedigree MatrixOrigin (Chinese HTAP database company); no named founders publicly identified vendor
- Memory volume / scale no scale claim vendor
- Performance 88.78% LMES (MatrixOrigin benchmark blog, 2025) vendor
- MCP support yes — native MCP server (memoria mcp) for any MCP-compatible agent repo
- A2A support not supported vendor
- Import / export MatrixOne SQL dump vendor
- Integration count 0 repo
- Orchestration agnostic repo
- Programmatic control full CRUD via lib repo
- Vendor benchmarks LMES 88.78% overall (vendor blog) vendor
- Validated verticals agent developers (horizontal); China market vendor
- Optimised for retrieval quality ("retrieval-strong") vendor
- Adjacent infrastructure BYO LLM + MatrixOne HTAP database (vector + graph + KV in one) vendor
- Pros Backed by MatrixOrigin's HTAP DB engine — memory benefits from mature transactional subs… vendor
- Cons Tied to MatrixOrigin database; primarily Chinese-market positioning. vendor
- Project & sources benchmark writeup vendor
- architecture-deployment-locus hybrid repo
- licensing-bucket open-source repo
- positioning-platform-vs-product platform repo
- taxonomy-three-layer-rollup agent-infra repo
- developer-activity-opendigger {"activity":7.32,"as_of":"2026-05","openrank_12mo":5.2,"openrank_current":2.69,"particip… third-party
- sdk-languages rest repo
- protocol-support mcp-server repo
What changed
Changes to this record's lifecycle, licence, release, pricing and hosting facts, from the catalog's git history. Subscribe (Atom)
- 2026-10-22
- License “not applicable — not OSS” → “Apache-2.0” source
- Latest release “not applicable — not OSS” → “v0.5.2 (Sep 11; page shows no year, presumably 2026)” source
- Deployment “Both (Memoria Cloud API + self-hosted MatrixOne DB)” → “Both — Memoria Cloud (thememoria.ai, remote mode) or self-hosted Docker Compose with MatrixOne (embedded mode); Kubernetes documented” source
- MCP support first recorded as “yes — native MCP server (memoria mcp) for any MCP-compatible agent” source