cognithor vs Memary
Local AI agent operating system vs. agent memory layer
cognithor is a local-first Python agent system that connects models, tools, and interfaces to plan and execute tasks. Memary is a Python-based memory layer with a knowledge graph and an included ReAct chat agent; its focus is retrieving personalized context for agents rather than providing a broader agent operating system.
cognithor: Run a Local Autonomous AI Agent
Cognithor is a Python agent operating system for running an assistant locally, connecting it to LLMs, messaging channels, memory, and tools. It suits users who want configurable personal automation and control over where their data is stored.
Memary: Add Memory and Knowledge Graphs to AI Agents
Memary is a memory layer for autonomous agents that combines a knowledge graph with user-focused memory to inform responses over time. It suits developers building personalized agents who can manage local models, database connections, and API credentials.
| cognithor | Memary | |
|---|---|---|
| Language | Python | Jupyter Notebook |
| License | Apache-2.0 | MIT |
| Stars | 158 | 2.7k |
| Forks | 25 | 206 |
| Last analyzed | Oct 3, 2026 | Oct 3, 2026 |
Key differences
- cognithor covers planning, policy checks, and tool execution, while Memary focuses on graph-backed memory, retrieval, and user context.
- cognithor supports local and cloud model providers, messaging and other interfaces, and tool integrations; Memary includes a ReAct agent, graph database options, and customizable tools.
- cognithor uses a six-tier memory system with hybrid search, while Memary tracks entities and their frequency and recency in graph-backed memory.
- cognithor is written in Python and licensed under Apache-2.0; Memary is listed as Jupyter Notebook and licensed under MIT.
- cognithor describes itself as pre-v1.0 beta and experimental; Memary describes its included agent as a demo-oriented implementation intended to be replaced.
- cognithor emphasizes local operation and configurable personal automation; Memary emphasizes adding personalized memory and knowledge-graph retrieval to agent workflows.
Choose cognithor if you…
- want a local-first assistant that can plan tasks and act through tools and multiple interfaces.
- need policy checks and approval controls before planned actions execute.
- want to prototype workflows across local and hosted models in a broader agent system.
Choose Memary if you…
- need graph-backed memory that tracks entities across user interactions.
- want to add multi-hop retrieval and personalized context to an agent workflow.
- plan to extend the included ReAct agent with custom tools and separate user contexts.
Compare the numbers side by side, or add more projects
This comparison is generated with AI from the OSRepos analyses of both projects. Always check each project's repository and documentation before choosing.