About Jonex
Jonex is an enterprise knowledge engine that turns multimodal company data into governed, source linked knowledge for AI assistants and agents. It ingests documents, spreadsheets, scans, images, audio, video, web pages and CAD files through one parsing path, compiles business entities, relationships, rules and evidence ahead of any query, and returns answers that keep a pointer back to the source document, version and location.
The product is organized around three engines: unified multimodal parsing, ontology based knowledge compilation, and a navigable wiki compiled from source material with hierarchical pages, linked concepts and backlinks. Teams can start from domain templates, connect sources such as file upload, REST API pulls, signed OpenAPI pushes, S3 and MinIO storage, NAS shares and cloud drives, and review compiled entities and citations in a search and verification view. Agents connect through an MCP server or a REST API. The vendor states that Jonex is a knowledge layer rather than an agent orchestration or workflow platform, and it lists use cases in finance, government, manufacturing, education and engineering.
The site is available in Chinese and English. The vendor reports benchmark results on HotpotQA, MuSiQue and 2WikiMHQA and a FinanceBench evidence tracing test, all on its own site.
The vendor does not publish prices. Access is through a request a trial form.