Categories
Developer Tools RAG AI Infrastructure Data Extraction Knowledge Graph Document Intelligence PDF ParsingStars
Forks
Watchers
Developer links
Knowhere
With 2,600+ GitHub stars since its May 2026 open-source launch, Knowhere solves the last-mile problem of document intelligence for AI systems — transforming complex unstructured PDFs, reports, and multi-page documents into structured JSON chunks that LLMs can consume without hallucination. The platform processes documents through an AI-native parsing pipeline that handles 20+ page documents with deep hierarchies, intricate tables, and multimodal content including images with OCR, achieving 95% precision in information extraction while reducing token costs by 50% compared to raw document ingestion. The knowledge tree architecture maintains historical context across multiple documents, enabling cross-document graph navigation for agentic retrieval that goes beyond simple chunk-based RAG. Built on Python 3.11+ with MinerU as the default PDF parser, the backend API runs alongside async workers that process document ingestion, graph construction, and embedding generation. The self-hosted Docker Compose stack packages the API server, processing workers, and Next.js dashboard for managing API keys, webhooks, and document-processing jobs, backed by PostgreSQL and Redis. Both Python and Node.js SDKs provide programmatic access for integration into existing AI pipelines and agent frameworks. LLM providers include DeepSeek and Alibaba Cloud DashScope with configurable key rotation for rate-limit management. Deploy on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache 2.0 licensed.
Benefits
- 95% Extraction Precision on Complex Docs
- Processes multi-page PDFs with deep hierarchies, nested tables, and multimodal content including image OCR, delivering structured JSON that eliminates hallucinations caused by broken text in downstream LLM reasoning.
- 50% Token Cost Reduction
- Knowledge tree architecture compresses and structures extracted content into optimized chunks, cutting token consumption in half compared to raw document ingestion while preserving full semantic context for accurate retrieval.
- Cross-Document Knowledge Graphs
- Builds graph relationships across multiple documents enabling agentic retrieval that navigates between related concepts, sections, and sources rather than isolated chunk-based similarity matching.
- Self-Hosted Document Intelligence
- Full Docker Compose stack keeps all documents, embeddings, and extracted knowledge on your infrastructure. No data leaves your environment except optional LLM API calls to your chosen provider.
Features
- AI-Native Document Parsing
- MinerU-powered pipeline extracts text, tables, images, and hierarchical structure from complex PDFs and documents, outputting structured JSON chunks ready for RAG and agent consumption.
- Knowledge Tree Architecture
- Maintains hierarchical context and cross-document relationships in a graph structure, enabling multi-hop retrieval that preserves section paths, parent-child relationships, and source attribution.
- Dashboard and API Management
- Next.js web interface manages API keys, monitors document-processing jobs, configures webhooks for completion notifications, and provides usage analytics for the self-hosted deployment.
- Multi-Provider LLM Support
- Configurable LLM backends including DeepSeek and Alibaba Cloud DashScope with comma-separated key pools for automatic rotation across provider quotas and rate limits.
- Python and Node.js SDKs
- Official SDKs for both Python and Node.js provide typed interfaces for document upload, job management, retrieval queries, and webhook configuration in existing agent pipelines.