Preview. These maps are our reading of public sources and haven't been checked with the projects' maintainers yet.
Mem0
Memory layer for agents and apps: library, self-hosted server, and hosted Mem0 Platform. LLM extracts facts from conversations into a vector store with entity linking.
mem0.ai/Active
mem0ai 2.2.1 on PyPI 2026-09-25; repo pushed 2026-10-23. New 'ADD-only' memory algorithm April 2026; OpenMemory removed from the monorepo 2026-07-29. (checked 2026-10-25)
In Agent memory
Where it stands apart
- Can it run entirely on your own device or infrastructure, without the maker's servers? partly (some features need the maker's servers) (14% of 21 others)
- By default, how do old or wrong memories mainly go away? they stay but are marked superseded (old facts are kept) (16% of 19 others)
- What is the main storage model for long-term memory? vector store of extracted facts or chunks (20% of 20 others)
What it shares with most of the community
- Is it actively maintained? active (67% of 21 others)
- How is the main thing licensed? permissive open source (MIT/Apache/BSD) (67% of 21 others)
- How is it mainly sustained? company with a paid product, tier, or ads (75% of 20 others)
- Who is it for? developers (76% of 21 others)
- In the default setup, who mainly decides what gets remembered? the system, by extracting memories from conversations or data in the background (67% of 21 others)
- How can the person the memories are about see and edit them? only through developer APIs or a developer dashboard (62% of 21 others)
- Does it work with models from different LLM vendors? yes, model-agnostic (86% of 21 others)
- Is there a first-party MCP server for it? yes (67% of 21 others)
- How is memory separated between users or tenants? built-in scoping by user, agent, session, or tenant IDs (62% of 21 others)
- Does it publish scores on public memory benchmarks (LoCoMo, LongMemEval, BEAM, etc.)? yes (62% of 21 others)
Nearest neighbors
- Memori 81% of 16 shared answers match
- Cognee 76% of 17 shared answers match
- Graphiti 76% of 17 shared answers match
Every answer
| Question | Answer | Source |
|---|---|---|
| Is it actively maintained? | active | source 2026-10-25 |
| How is the main thing licensed? | permissive open source (MIT/Apache/BSD) Apache-2.0 | source 2026-10-25 |
| How is it mainly sustained? | company with a paid product, tier, or ads | source 2026-10-25 |
| Can it run entirely on your own device or infrastructure, without the maker's servers? | stands apart partly (some features need the maker's servers) Library and self-hosted server run locally; README table says advanced features are 'Teasers' self-hosted and 'All included' on the Cloud Platform; published benchmark scores reflect platform-only optimizations. | source 2026-10-25 |
| Do you need an account to use it? | optional account OSS library needs no Mem0 account (needs an LLM key); hosted Mem0 Platform account optional. | source 2026-10-26 |
| Who is it for? | developers | source 2026-10-25 |
| Which shared standard or format does it mainly build on? | none of these (its own format or interface) Own API over pluggable vector stores; MCP one interface. | source 2026-10-26 |
| In the default setup, who mainly decides what gets remembered? | the system, by extracting memories from conversations or data in the background add() runs LLM extraction over messages; via MCP the agent decides when to save. | source 2026-10-25 |
| How can the person the memories are about see and edit them? | only through developer APIs or a developer dashboard Developer dashboard (cloud and self-hosted server) and APIs. | source 2026-10-25 |
| By default, how do old or wrong memories mainly go away? | stands apart they stay but are marked superseded (old facts are kept) April 2026 algorithm is 'single-pass ADD-only extraction ... no UPDATE/DELETE. Memories accumulate; nothing is overwritten', with temporal ranking of the current instance. Explicit delete APIs still exist. | source 2026-10-25 |
| Where do memories live in the default setup the docs lead with? | in the developer's own infrastructure README leads with the library (local), then self-hosted, then cloud; the MCP server is hosted only. | source 2026-10-25 |
| Does it work with models from different LLM vendors? | yes, model-agnostic | source 2026-10-25 |
| What is the main storage model for long-term memory? | stands apart vector store of extracted facts or chunks Vector store plus BM25 and entity linking. | source 2026-10-25 |
| What shape does it mainly take (the one its docs lead with)? | library or API primitive you build into your app Also a self-hosted server and hosted platform. | source 2026-10-25 |
| Is there a first-party MCP server for it? | yes Hosted Mem0 MCP server; 'Nothing runs on your machine'. | source 2026-10-25 |
| How is memory separated between users or tenants? | built-in scoping by user, agent, session, or tenant IDs user_id, agent_id, run/session IDs. | source 2026-10-25 |
| Does it publish scores on public memory benchmarks (LoCoMo, LongMemEval, BEAM, etc.)? | yes LoCoMo 92.5, LongMemEval 94.4, BEAM (platform numbers). | source 2026-10-25 |
Sources
This page was derived from:
- GitHub · github.com/mem0ai/mem0 · read 2026-10-25
- Web page · docs.mem0.ai/platform/mem0-mcp · read 2026-10-25
- Web page · mem0.ai/ · read 2026-10-25
Are you a maintainer? Claiming and correcting projects is coming soon.