WeKnora is an open-source LLM knowledge platform that transforms raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki, addressing the challenge of managing and querying large document collections efficiently.
Source: Official README and linked documentation View on GitHub →WeKnora is being discussed because its coverage to document understanding and knowledge management, addressing the pain points of inefficient knowledge retrieval and manual document organization. Its unique technical choices include a modular architecture, support for multiple LLM providers, and integration with various data sources and IM channels.
Source: Official README and linked documentationEnables fast and accurate answers from knowledge bases using Retrieval-Augmented Generation (RAG) techniques.
Source: Official README and linked documentationAn autonomous reasoning agent that orchestrates complex tasks involving retrieval, MCP tools, and web search.
Source: Official README and linked documentationAutomatically generates structured, interlinked markdown Wiki pages from raw documents, creating a self-maintaining knowledge base.
Source: Official README and linked documentationThe architecture is modular, allowing for the swapping and extension of components such as document parsing, vectorization, retrieval, and LLM inference. It supports local/private cloud deployment and features a zero-barrier Web UI for onboarding.
Source: Official repository tree and dependency filesgin-gonic/gingo-sql-driver/mysqlgo-viper/mapstructure/v2github.com/elastic/go-elasticsearch/v7github.com/elastic/go-elasticsearch/v8WeKnora is suitable for enterprises dealing with large document collections, such as knowledge management, customer support, and internal knowledge sharing. It is useful in scenarios where efficient knowledge retrieval and automated document organization are critical, such as creating a company-wide knowledge base or a customer support knowledge base.
Source: Official README and linked documentationAs of 2026-09-23, the latest release is v0.8.0, published on 2026-09-03. It adds a skill sandbox, tenant skill catalog, cross-session memory, anydoc, GitLab and Tencent IMA sources, LiteLLM, more search services, OIDC, and security hardening.
Source: github.com/Tencent/WeKnora/releases/tag/v0.8.0status: docs-only | checked_at: 2026-09-23 | environment: Docs only. Reviewed the official README, CHANGELOG, LICENSE,.env.example, docker-compose.yml, and GitHub Releases. | limitations: No image was pulled. Frontend, backend, database, docreader, vector database, Langfuse, and sandbox were not started. No document was imported, and no model, data source, or IM credential was configured. | observed_output: The latest release is v0.8.0, published on 2026-09-03. Core startup requires Docker, Compose, and Git. The Web UI defaults to port 80 and the API to 8080. The license says the project core is MIT while third-party components retain their own terms.
Source: github.com/Tencent/WeKnora/releases/tag/v0.8.0RAGFlow also provides document parsing, retrieval-augmented generation, agents, and self-hosting, with an official emphasis on visual document understanding, RAG workflows, and agents. WeKnora also includes a maintained wiki, page revisions, a tenant skill catalog, a session sandbox, and cross-session memory. Start with RAGFlow for complex document parsing and RAG workflows. Evaluate WeKnora when the wiki and skill sandbox need to work together.
Source: github.com/infiniflow/ragflowWeKnora is a project for organizations seeking to efficiently manage and query large document collections. Its broad features and modular architecture make it suitable for a wide range of use cases, particularly in enterprise environments where knowledge management is crucial.
Source: Official repositoryThe latest release is v0.8.0, published on 2026-09-03. Core startup requires Docker, Compose, and Git. The Web UI defaults to port 80 and the API to 8080. The license says the project core is MIT while third-party components retain their own terms. No image was pulled.
The review environment was: Docs only. Reviewed the official README, CHANGELOG, LICENSE,.env.example, docker-compose.yml, and GitHub Releases. WeKnora behavior outside that scope still needs a local check.