# EdgeQuake

> Apache-2.0 GraphRAG engine in Rust, runs on local models

**Category:** AI Agents  
**Pricing:** open-source  
**URL:** https://github.com/raphaelmansuy/edgequake  
**GitHub:** https://github.com/raphaelmansuy/edgequake  
**Added:** 2026-10-14

---

## The Hook

EdgeQuake turns documents into a graph of entities and relationships, then answers questions by combining vector search with graph traversal — the LightRAG approach, implemented in Rust.

## What's Good

Apache-2.0 (raphaelmansuy/edgequake), release v0.35.0. A setup wizard detects local model servers — Ollama, LM Studio, MLX or llama.cpp — so it can run fully on your own machine with no API key; OpenAI or Anthropic keys are optional for faster results. Binds to localhost by default. ★~2.2k.

## The Catch

Pre-1.0 with a small group of contributors (about a dozen), so expect breaking changes. Graph quality depends heavily on the model you use for entity extraction — small local models produce thinner, noisier graphs, and the first local run pulls several GB of models. Vs cognee (a more established memory and knowledge-graph layer for agents): EdgeQuake is the lighter Rust option for local GraphRAG.

## Verdict

Local-first GraphRAG in Rust — worth a try for private documents, if you can give it a capable extraction model.