nano-graphrag and graphiti
The projects are competitors, as both offer GraphRAG implementations, with Graphiti appearing to be a more comprehensive knowledge graph solution and nano-graphrag a simpler, more hackable implementation.
About nano-graphrag
gusye1234/nano-graphrag
A simple, easy-to-hack GraphRAG implementation
Builds knowledge graphs from text by extracting entities and relationships, then performs retrieval-augmented generation through both global and local graph traversal modes. Supports pluggable components including multiple LLM providers (OpenAI, Bedrock, Ollama), vector databases (FAISS, Milvus, HNSWlib), and graph backends (Neo4j, NetworkX), with full async/await support and MD5-based deduplication for incremental inserts.
About graphiti
getzep/graphiti
Build Real-Time Knowledge Graphs for AI Agents
Provides temporal validity windows and provenance tracking for facts, enabling historical queries and full lineage from derived data to source episodes. Supports both prescribed (Pydantic-defined) and learned ontologies, with incremental graph updates that avoid costly recomputation. Integrates hybrid retrieval combining semantic embeddings, BM25 keyword search, and graph traversal for low-latency context assembly without LLM summarization.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work