RAG и базы знаний
Local_Pdf_Chat_RAG
Transparent Python RAG reference with FAISS + BM25 hybrid retrieval.
★ 957 звёздMIT
Смотреть руководство →ragent — проект с открытым исходным кодом из категории «RAG и базы знаний» от nageoffer (nageoffer/ragent на GitHub). У него 4 208 звёзд на GitHub и 848 форков, он распространяется по лицензии Apache-2.0 и последний раз обновлялся на GitHub 7 октября 2026.
Ragent (容纳智能体): a production-grade Java Agentic RAG platform with hybrid retrieval and an admin console
Read the overview of the seven Maven modules and the infrastructure you will need (database, Redis, vector store).
Get the backend and console source from GitHub.
Build and run the platform using the step-by-step guide on the docs site, configuring your LLM providers.
Use the admin console to ingest documents into a knowledge base and run your first cited query.
Vector, keyword, knowledge-graph, and web search run together, then dedup, RRF fusion, and reranking produce the final context.
Query mapping, rewriting, and splitting, plus a tree-shaped intent recognizer that routes across multiple knowledge bases.
Multiple model tiers with first-token probing and circuit-breaker fallback when a provider degrades.
AgentScope-based discovery and calling of MCP tools as part of answering.
Redis fair queueing and distributed concurrency limits keep shared deployments stable under load.
A React admin console, answer sourcing/citations, feedback capture, and Langfuse tracing.
Clone the project and enter its directory.
git clone https://github.com/nageoffer/ragent.git
cd ragentEnvironment setup, configuration, and run commands are maintained on the project's documentation site.
# see the quick-start guide at nageoffer.com/ragent for the exact commandsStart the backend and console per the guide, then sign in to ingest documents and test queries.
# see the quick-start guide for service URLs and default credentialsDeveloper nageoffer maintains Ragent and its documentation site.
Yes, Apache-2.0 licensed; you run it on your own infrastructure and supply LLM provider keys.
It needs LLM (and typically embedding) provider access configured per the quick-start guide; costs depend on the providers you choose.
Those are broad low-code platforms; Ragent is a code-first Java/Spring platform focused on retrieval quality, model resilience, and production traffic controls that Java teams can extend directly.
Clone the repo and follow the quick-start guide at nageoffer.com/ragent to configure providers, start the services, and ingest your first knowledge base.
Transparent Python RAG reference with FAISS + BM25 hybrid retrieval.
Rag.computer — self-hostable rag platform - document ingestion, embedd, open-source.
AI-native data platform: 60+ connectors, ETL, scheduling, and RAG search.
Local AI-driven document search and editing with first-in-class hybrid retrieval
Vespa — the AI search platform: vector, text and tensor search at production scale.
All-in-one API platform for search, recommendations, RAG and analytics — self-hostable
Другие проекты из категории «RAG и базы знаний» с наибольшим числом звёзд в каталоге LibreAI — реалистичный список альтернатив, если ragent вам не подходит.
Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or…
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context…
Lightweight OCR toolkit turning PDFs and images into structured, LLM-ready data in 100+ languages
MinerU — turns PDFs and Office documents into LLM-ready Markdown and JSON for agents and RAG pipelines, Apache-2.0 plus terms
Stop renting your intelligence. Own it with AnythingLLM. Everything you need for a powerful local-first agent experience
Соседние проекты из категории «RAG и базы знаний» в каталоге, ближайшие по числу звёзд к ragent — продолжайте изучать полку по обе стороны от него.
RAGApp: deploy a RAG app over your documents with Docker — configure it like a custom GPT
nano-graphrag: a small, hackable GraphRAG implementation in about 1,100 lines — no Microsoft GraphRAG bulk
AI-native database for hybrid search across dense, sparse, and tensor vectors plus full text
The open-source, permission-aware context layer that lets agents search all your company knowledge.
OpenKB compiles your documents into a persistent, interlinked wiki knowledge base — no vector DB
Knowhere turns documents into structured, hierarchy-native chunks ready for agents and RAG.
Бесплатно и открыто — забирайте исходный код, ставьте звезду репозиторию и начинайте создавать.