Freedom251 › Retrieval-Augmented Generation Tools
Graphiti
What Graphiti is
Graphiti is an open-source framework for building and querying temporal context graphs for AI agents. It records when facts become valid and when they are superseded, allowing an application to retrieve current information or ask what was true at an earlier point. New structured and unstructured data can be incorporated incrementally rather than requiring a complete graph rebuild. Graphiti links derived entities and relationships to the source episodes they came from, preserving provenance for retrieved information. Its hybrid retrieval brings together semantic search, keyword search, and graph traversal, with result reranking available. Developers can define entity and edge types with Pydantic models, or allow graph structure to emerge from data. Supported graph backends listed by the project include Neo4j, FalkorDB, and Amazon Neptune; Kuzu is deprecated. Graphiti defaults to OpenAI and also supports Anthropic, Google Gemini, Groq, and compatible endpoints for other OpenAI-compatible providers. The project includes a FastAPI REST service and an MCP server for episode and entity management, relationships, search, groups, and graph maintenance. It can run locally and is self-hosted and self-managed. Setup requires Python 3.10 or higher and an LLM API key; structured-output-capable LLM services work best. It is Apache-2.0 licensed and costs nothing.
Who it suits
Graphiti suits developers building AI-agent systems that need time-aware facts, source provenance, and hybrid graph retrieval. It is aimed at teams prepared to host and operate the framework and its surrounding system themselves.
What it does well
- Queries can retrieve facts from a chosen point in time.
- Incremental updates avoid complete graph recomputation.
- Hybrid retrieval combines semantic, keyword, and graph search.
- Pydantic models support developer-defined entity and edge types.
- Includes both MCP server and FastAPI REST service.
- Apache-2.0 licensed and available for free.
What to know first
- Requires Python 3.10 or higher and an LLM API key.
- Self-hosting means users operate the surrounding system.
- Enterprise SLAs, support, and security guarantees belong to Zep.
- Kuzu is marked deprecated.
Limits
The open-source framework is self-managed and self-hosted, so users are responsible for building and operating the surrounding system. The listed price is free under the Apache-2.0 license.
Plans
| Plan | Billed and limits | Price |
|---|---|---|
| Graphiti (open-source) | Apache-2.0 licensed framework · self-managed | Free |
As github.com lists them, read 9 Oct 2026. The maker's own page has the final word.
Getting started
- Open the Graphiti project on GitHub.
- Prepare Python 3.10 or higher and an LLM API key.
- Choose a supported graph backend: Neo4j, FalkorDB, or Amazon Neptune.
- Select an LLM provider; Graphiti defaults to OpenAI and supports Anthropic, Google Gemini, Groq, and compatible endpoints.
- Run Graphiti locally or deploy it in a self-hosted environment.
Verdict
Choose Graphiti if you need an open-source, self-managed framework for temporal context graphs, with provenance and multiple retrieval methods. Look elsewhere if you need enterprise SLAs, support, or security guarantees as part of the offering; those are features of Zep.
Questions
How much does Graphiti cost?
The listed plan is free. Graphiti is open source under the Apache-2.0 license.
Which platforms does Graphiti support?
Its listed platforms are API and self-hosted. The project describes it as runnable locally and includes a FastAPI REST service.
Which graph databases can it use?
The project lists Neo4j, FalkorDB, and Amazon Neptune as supported graph backends. Kuzu is marked deprecated.
Which LLM providers are supported?
Graphiti defaults to OpenAI and also supports Anthropic, Google Gemini, and Groq. OpenAI-compatible providers can be used through compatible endpoints.
What does setup require?
The listed requirements include Python 3.10 or higher and an LLM API key. Structured-output-capable LLM services work best.
Is managed support included?
Graphiti is self-managed and self-hosted. Enterprise SLAs, support, and security guarantees are features of Zep.
Facts from Graphiti's own pages
- API
- The repository describes a REST API service built with FastAPI. · github.com, 4 Oct 2026
- Changing facts
- It tracks when facts become valid and when they are superseded, preserving historical information for point-in-time queries. · github.com, 4 Oct 2026
- Custom ontology
- Developers can define entity and edge types with Pydantic models or let structure emerge from data. · github.com, 9 Oct 2026
- Data and ontology
- It incrementally integrates structured and unstructured data and supports developer-defined entity and edge types using Pydantic models. · github.com, 4 Oct 2026
- Database integrations
- The README lists Neo4j, FalkorDB, and Amazon Neptune as supported graph backends and marks Kuzu deprecated. · github.com, 9 Oct 2026
- Deployment
- The maker’s Graphiti page describes the framework as runnable locally and distinguishes it from Zep’s managed commercial platform. · getzep.com, 4 Oct 2026
- Deployment and support
- The README describes Graphiti as self-hosted and self-managed, with users responsible for building and operating the surrounding system. · github.com, 9 Oct 2026
- Graph databases
- The repository lists Neo4j, FalkorDB, and Amazon Neptune as supported graph backends, and marks Kuzu as deprecated. · github.com, 4 Oct 2026
- Graph updates
- New data is integrated incrementally without requiring complete graph recomputation. · github.com, 9 Oct 2026
- License
- The GitHub repository identifies the project as Apache-2.0 licensed. · github.com, 9 Oct 2026
- LLM integrations
- Graphiti defaults to OpenAI and also documents Anthropic, Google Gemini, Groq, Azure OpenAI, and OpenAI-compatible providers. · github.com, 9 Oct 2026
- MCP
- The repository includes an MCP server for episode and entity management, relationship handling, semantic and hybrid search, group management, and graph maintenance. · github.com, 4 Oct 2026
- MCP and REST
- The repository includes an MCP server for assistant integrations and a FastAPI REST service. · github.com, 9 Oct 2026
- Provenance
- Entities and relationships trace back to the raw episodes that produced them. · github.com, 9 Oct 2026
- Purpose
- Graphiti is an open-source framework for building and querying temporal context graphs for AI agents. · github.com, 9 Oct 2026
- Requirements
- The README lists Python 3.10 or higher and an LLM API key among its requirements. · github.com, 9 Oct 2026
- Retrieval
- Hybrid retrieval combines semantic embeddings, BM25 keyword search, and graph traversal. · github.com, 9 Oct 2026
- Security and support
- The maker’s comparison says Graphiti is self-managed and self-hosted, while enterprise SLAs, support, and security guarantees are features of Zep. · github.com, 4 Oct 2026
- Security reporting
- The security policy directs vulnerability reports through GitHub Private Vulnerability Reporting and lists version 0.x as supported for security updates. · github.com, 9 Oct 2026
- Telemetry
- Graphiti collects anonymous usage statistics, and its README says telemetry can be disabled with GRAPHITI_TELEMETRY_ENABLED=false. · github.com, 9 Oct 2026
- Telemetry data
- The README says telemetry does not collect API keys, graph content, queries, or personal identifiers. · github.com, 9 Oct 2026
- Temporal facts
- Facts have validity windows, so users can query what is true now or what was true at a point in the past. · github.com, 9 Oct 2026
- What it does
- Graphiti is an open-source framework for building and querying temporal context graphs for AI agents. · github.com, 4 Oct 2026