agentmemory vs cognee
The two are close in size: 28.0k stars for agentmemory, 30.5k for cognee. Over the days we have tracked them agentmemory moved +25.3% and cognee +19.7%, so agentmemory is growing faster right now.
They split the axes: agentmemory leads on setup ease, cognee on context depth.
Stars and commit dates come from our own daily tracking. The six axes are read off each project's documentation by our review pipeline, so they describe what a project says about itself, not what we measured in its code.
Where they stand today
Provides a local, multi-agent persistent memory server with hybrid retrieval and confidence scoring that shares memories across any MCP/HTTP agent.
- Stars
- 28.0k
- Tracked growth
- +25.3%
- Maturity
- ●●●●●
- Last commit
- 5d ago
- Language
- TypeScript
- License
- Apache-2.0
- Cost to run
- Free to run locally; optional cloud API keys for agent LLMs.
Runs the full memory layer (graph, vectors, sessions) on a single self-hosted Postgres-backed stack, combining vector search and evolving knowledge graphs.
- Stars
- 30.5k
- Tracked growth
- +19.7%
- Maturity
- ●●●●●
- Last commit
- 54m ago
- Language
- Python
- License
- Apache-2.0
- Cost to run
- Your LLM API key, or use Cognee Cloud (managed)
Six axes, head to head
Each axis runs 0 to 5. The label under a score is what that project's own docs claim, not a category average.
| Axis | agentmemory | cognee |
|---|---|---|
Context depth How much of your codebase it sees before it answers: the open diff, the diff plus related files, or the whole repository. | ●●●●● Diff + related files | ●●●●● Whole-graph memory |
Noise control How it keeps output volume down — severity thresholds, deduplication, incremental runs over new commits only. | ●●●●● Confidence scoring | ●●●●● Feedback & routing |
Customization How far it bends to your team: custom rules, prompts, style guides, per-path config. | ●●●●● Config file options | ●●●●● Config + plugins |
Privacy Whether your code stays on your own infrastructure: fully local, self-hostable, or cloud API only. | ●●●●● Fully local | ●●●●● Self-hostable |
Model freedom Whether you can point it at any provider, or it is wired to one. | ●●●●● Provider-agnostic | ●●●●● Bring-your-key |
Setup ease What it takes to get a first useful run out of it. | ●●●●● One-command setup | ●●●●● API key required |
Which one to pick
Pick agentmemory if…
Easy setup — one-command local memory server with a demo and adapters for many coding agents, preserving session memory across restarts.
- Setup ease: One-command setup (5/5 against 3/5)
Pick cognee if…
Self-hostable — run a unified, persistent agent memory on a single Postgres instance with minimal extra services and strong integration options.
- Context depth: Whole-graph memory (5/5 against 3/5)
What people want from each one
rohitg00/agentmemory
topoteretes/cognee
Hacker News: Show HN: Cognee – Open-Source AI Memory Layer That Remembers Context drew 9 points and 2 comments.
- [Hackathon] Vellum integration — cognee memory nodes for Vellum Workflows 👍 0
- [Bug]: recall ratifies affirmative/keyword queries as stored facts — and can harden them (e.g. 'added to watch list' becomes 'merged') 👍 0
- [Bug] GET /datasets/{id}/data and /data/{id}/raw return 500 (IndexError) instead of 404 when dataset is not found 👍 0
Questions people ask
Is agentmemory better than cognee?
They split the axes: agentmemory leads on setup ease, cognee on context depth. agentmemory is worth picking when easy setup — one-command local memory server with a demo and adapters for many coding agents, preserving session memory across restarts.
Which of agentmemory and cognee keeps my code private?
agentmemory: Fully local (5/5). cognee: Self-hostable (4/5).
What does each one cost to run?
agentmemory: Free to run locally; optional cloud API keys for agent LLMs.. cognee: Your LLM API key, or use Cognee Cloud (managed).
Full profiles: rohitg00/agentmemory and topoteretes/cognee. Everything else in Memory & context.