WeKnora
VerifiedOfficialby Tencent
Official listing: built and published by Tencent. More from Tencent
Turns documents into a queryable knowledge base with RAG question answering, a ReAct agent orchestrating retrieval, MCP tools and web search, and an auto-generated wiki.
- Access
- APIMCPCLIExtension
- Auth
- API Key
- Skills
- 5
Skills
Document Ingestion
Parses PDF, Word, Excel, PPT, EPUB, images and more, with per-upload parser, chunking, OCR and VLM settings plus batch reparse.
Hybrid Retrieval
Retrieves passages via BM25, dense vectors, GraphRAG and parent-child chunking across pgvector, Elasticsearch, Milvus and Qdrant.
ReAct Reasoning Agent
Runs progressive multi-step ReAct reasoning that autonomously chains knowledge retrieval, MCP tools and web search to reach an answer.
Auto Wiki Generation
Distills raw documents into structured interlinked Markdown wiki pages with an interactive knowledge graph browser in the UI.
Data Source Sync
Auto-syncs knowledge from Feishu, Notion, Yuque and RSS with incremental or full sync, and serves answers through IM and widgets.
Related Agents
LightRAG
by HKU Data Science Lab
Builds a graph-based retrieval index over documents so agents can answer questions spanning many sources, served…
FastGPT
by labring
Builds AI agents on a visual Flow canvas backed by hybrid-retrieval knowledge bases, workflow and RPA nodes…
RAGFlow
by InfiniFlow
Open-source RAG engine with deep document parsing (tables, charts, images) and an official MCP server. Deploy via…
MinerU
by OpenDataLab
Converts PDFs, Office documents, and scanned pages into LLM-ready Markdown or JSON, preserving tables, formulas, and…