Connected memory engine
Combines vector search, graph relationships, and relational storage into one memory layer for connected retrieval.
Cognee
Freemium
Cognee is an open-source memory platform for AI agents and applications. It connects data across sources, builds connected memory using vector, graph, and relational storage, and makes that context available through SDK, cloud, or self-hosted deployments.
What is Cognee?
Cognee is an open-source memory platform for AI agents and applications. It ingests information from sources such as Slack, GitHub, Linear, documents, warehouses, and APIs, then organizes it into connected, queryable memory. The platform combines vector search, knowledge graphs, and relational storage so agents can retrieve relationships and context across sources rather than only similar text fragments.
Cognee can be used through its SDK, Cognee Cloud, or a self-hosted deployment. It integrates with agent tools and frameworks including Claude Code, Cursor, Codex, LangGraph, n8n, CrewAI, the OpenAI Agents SDK, Google ADK, and MCP-compatible clients. Cloud provides multi-tenant workspaces and per-session memory, while self-hosting supports teams that need more control over infrastructure and data boundaries.
What can Cognee do?
Data ingestion and connectors
Brings together sources such as Slack, Notion, Google Drive, GitHub, Confluence, Jira, Dropbox, Amazon S3, and Salesforce, with additional data and API connections supported through the platform.
Generated or custom data models
Builds knowledge graphs with automatically generated data models, while allowing teams to define custom ontologies and domain-specific schemas.
Agent integrations
Provides an SDK and MCP server, with integrations for Claude Code, Cursor, Codex, LangGraph, n8n, CrewAI, the OpenAI Agents SDK, Google ADK, and other compatible clients.
Deployment and operational controls
Runs locally, on a team’s own infrastructure, or through Cognee Cloud. Enterprise options include bring-your-own-cloud deployment, SLAs, dedicated support, provenance on answers, and personalization per user and agent.
Use Cases
“Persistent agent memory”
Give coding and task-oriented agents context that carries across sessions, so they can recall prior work instead of starting from an empty context each time.
“Internal company copilots”
Connect team documents, tickets, conversations, repositories, and other sources so employees can search a shared, connected knowledge layer.
“GraphRAG and semantic search”
Build retrieval systems that use both semantic similarity and relationships between entities for document intelligence, research, and knowledge-base applications.
“Customer and user support”
Enrich support agents with connected customer or product information and maintain per-user or per-session memory in multi-tenant applications.
“Domain-specific agents”
Define custom ontologies and deploy memory in a private or bring-your-own-cloud environment when agents must follow organization-specific rules or data boundaries.
Frequently Asked Questions
What is Cognee?
Cognee is an open-source memory platform for AI agents. It turns source data into connected, queryable memory using vector search, knowledge graphs, and relational storage.
How does Cognee process data?
It ingests data, generates a knowledge graph and data model, and stores the result across vector, graph, and relational layers. Teams can use automatically generated models or define custom models and ontologies.
Can Cognee be self-hosted?
Yes. Cognee can run locally or on a team’s own infrastructure. The site also describes air-gapped and bring-your-own-cloud deployment options for teams with stricter infrastructure or data-boundary requirements.
Which tools can connect to Cognee?
Cognee supports Claude Code, Cursor, Codex, LangGraph, n8n, CrewAI, the OpenAI Agents SDK, Google ADK, and MCP-compatible clients. Data connectors include Slack, Notion, Google Drive, GitHub, Confluence, Jira, Dropbox, Amazon S3, and Salesforce, among others.
How is Cognee priced?
Cognee Cloud has a free tier with one workspace and 1M included tokens, followed by usage-based Standard pricing at $1 per 1M tokens; additional workspaces cost $5 per month each. The open-source engine can be self-hosted for free, while Enterprise is offered as a bring-your-own-cloud engagement with support and service-level options.
Quick Facts
- Category
- AI infrastructure and developer tool
- Primary users
- Agent builders, product and platform teams, and enterprises
- Deployment
- Cognee Cloud, local, self-hosted, and bring-your-own-cloud
- Core storage model
- Vector, graph, and relational layers
- Notable workflow
- Ingest data, generate or customize a data model, then recall connected context through agents and applications
Cognee Alternatives
Statewave
www.statewave.ai
Statewave is an open-source, self-hosted memory runtime for AI agents. It stores durable episodes, compiles structured memories with provenance, and returns ranked context bundles for applications that need continuity across sessions.
Second Brain
www.thesecondbrain.dev
Second Brain gives connected AI tools a shared memory for projects, decisions, preferences, and updates. It runs in your own Cloudflare account and lets you save, recall, inspect, and organize context across supported clients and capture sources.
Firetower
usefiretower.com
Firetower is an open-source, self-hosted control plane for running coding agents on servers you own. It lets developers start agent sessions from repositories, monitor them from desktop or mobile clients, review changes, and ship branches without keeping the initiating device online.
iphone-use
github.com
Open-source, self-hosted iPhone control for AI agents on macOS.
Steel
steel.dev
Steel is open-source browser infrastructure for AI agents and web automation. It provides cloud browser sessions through an API, with support for common browser automation tools and workflows such as scraping, research, RPA, and agent-driven web tasks.
ByteRover
byterover.dev
ByteRover is a context layer for AI agents and teams, storing project knowledge as memory files with control over sharing across sessions, agents, and teammates. Enterprise options support on-premise and private-VPC deployment.