Skip to content
repository radar
RAG & Memory

🧠Graphiti

ACTIVE BREAKOUT

getzep/graphiti · homepage ↗

Build Real-Time Knowledge Graphs for AI Agents

⭐ Very popular: 30k stars, gaining about 246 a week

View on GitHub ↗

repo profile

vintage 2024 2 years old
delivery library
language Python
license Apache-2.0

momentum

total stars 30k
stars added last week +246
HN peak 142 pts 1y 11mo ago
weekly downloads 327k /wk · PyPI
used by 355 repos & packages
commits / week 13 accelerating · +279% vs prior mo
issues closed 32% ~6d to close, median
contributors 58 (+1 in 15d)
release cadence fortnightly
last activity today

durability

backing VC-backed Y Combinator
openness permissive
bus factor 2 concentrated
top-author share 37% 6 mo

bus factor = how many people it takes to cover more than half the commits (6 months). 1 is a solo project; higher means the work is spread across a team. top-author share is the single busiest author's slice of those commits.

since we covered it

monthly average + 1k/mo (+20%/mo) · + 23k total since PR#7

why it's a big deal

  • Graphiti maintains knowledge graphs that update as new data arrives, avoiding the batch recomputation and static summarization that standard RAG pipelines rely on.
  • It tracks when facts become true and when they are superseded, so agents can reason over data that changes over time rather than a single frozen snapshot.
  • It helps teams building AI agents that operate on evolving datasets and need queryable, source-traceable memory instead of re-embedding documents on every change.

under the hood

  • Written in Python, it stores entities as nodes and facts as temporal relationship edges with validity windows, plus episodes that retain the raw source data for provenance.
  • Retrieval combines semantic embeddings, BM25 keyword search, and graph traversal, returning results without depending on an LLM summarization step.
  • It runs on Neo4j, FalkorDB, or Amazon Neptune, defaults to OpenAI while supporting Anthropic, Google Gemini, Groq, and OpenAI-compatible local endpoints, and requires models with structured output support.

Radar summary, generated from the project's public sources

star history

PR#7 · 7k30k now Aug 2024Aug 2026
  1. PR#7 7k 2025-04-30
  2. now 30k + 23k since first covered

curve is sampled from GitHub's star history, plus our own daily readings since we covered it; the dashed stretch is before we first covered it, the solid line since. figures at coverage are the numbers we printed then (approx.), current count is live.

understory

Output and attention are roughly in balance.

-5 understory score output 68 · clout 73
Aug 2025 Jul 2026
  • output, commits & releases
  • clout, star velocity

output = commits & releases; clout = star velocity, both 0 to 100 monthly indices; the gap where output runs above clout is the understory. The understory →

covered in

  • PR#7 2025-04-30 below the radar

    Real-Time Knowledge Graphs for AI Agents

similar projects

compare these →
  • 🧠 Mem0

    2.1× the stars

    Universal memory layer for AI Agents

    63k ACTIVE
  • 📑 PageIndex

    📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG

    35k ACTIVE
  • 🖥️ Open WebUI

    5× the stars

    User-friendly AI Interface (Supports Ollama, OpenAI API,...)

    149k ACTIVE

comments

Sign in with GitHub to add your blip on Graphiti.

loading comments…