Enterprise AI search platform
Onyx is an open-core AI knowledge platform for workplace search, RAG chat, deep research, and internal assistant workflows. It connects to 40+ workplace applications, preserves document-permission context, and combines hybrid retrieval with agent features such as custom AI agents, Actions, MCP/OpenAPI integrations, web search, code interpreter support, and image-generation workflows. The product is strongest when a company needs an AI layer over many internal sources rather than a standalone chatbot.
Deployment and pricing
Deployment options include Onyx Cloud, Docker-oriented self-hosting, Kubernetes production paths, and enterprise self-hosted scenarios for organizations that need stronger governance or infrastructure control. Current public pricing lists Business at $20 per user per month with annual billing and Enterprise as custom. Infrastructure, model usage, support, and security requirements can change the effective cost for self-hosted or enterprise deployments.
License and due diligence
Onyx has substantial GitHub traction with more than 30K stars, but CMS copy should describe it as open-core rather than plain MIT across the whole product. The GitHub API reports NOASSERTION license metadata, and the repository license file places non-enterprise portions under MIT Expat while ee directories are covered by the Onyx Enterprise License. Buyers should verify license scope, connector behavior, and permission handling before rolling it out against sensitive company knowledge.
