#4040th of 42 in RAG, knowledge and search
kag
Knowledge-graph Q&A framework with logic-form reasoning, built on OpenSPG
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 9.1k
- License
- Apache-2.0
- Last commit
- Jan 2026
- Last release
- Jun 2025
- Language
- Python
8,536 on 3 Apr 20269,143 on 30 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
KAG builds a knowledge graph from structured and unstructured data on the OpenSPG engine, then answers multi-hop questions with planning, retrieval and reasoning steps guided by logical forms. Graph entities and source text chunks are indexed against each other. Version 0.8.0 adds MCP integration and several built-in index types. It ships as a web product and as a Python toolkit.
Who it is for: Teams building multi-hop Q&A over domain knowledge bases with knowledge graphs
Strengths
- Graph and text chunks are mutually indexed, so answers can link back to source passages
- Schema-constrained extraction supports domain-specific entity and event types
- Solver combines retrieval, graph reasoning, language reasoning and numeric calculation
- MCP support for public data sources and for use inside agent workflows
Weaknesses
- Requires the OpenSPG engine and its Docker Compose stack before any use
- The kag-model component is not yet open-sourced
- Detailed usage guides are hosted on Yuque, not in the repo
- Last release was 2025-06-28; RAM and GPU needs are not stated
What it needs
- Needs OpenSPG engine, Docker, Docker Compose, LLM provider
- Models: KAG-Thinker
- port 8887
Quick answers
Is kag open source?
Yes. Its code is on GitHub under the Apache-2.0 license.
Also in RAG, knowledge and search
See all 42| Rank | Project | Score |
|---|---|---|
| 1 | RAGFlowRAG engine with deep document parsing, template chunking and agentic retrieval | 90 out of 100 |
| 2 | LightRAGGraph-based RAG framework with a REST API server and web UI | 89 out of 100 |
| 3 | deeptutorSelf-hosted AI tutor with knowledge bases, guided learning, books and quizzes | 88 out of 100 |
| #4 | MaxKBEnterprise knowledge-base agent platform with RAG, workflows and MCP tools | 86 out of 100 |
| #5 | quivrGo engine that ingests content streams for hybrid search and keyword alerts | 84 out of 100 |