The No. 1 deep researcher beats Claude and ChatGPT with a trick neither uses.

Onyx is an innovative AI platform that serves as a knowledge and context layer for teams and AI agents. It connects to applications, indexing and surfacing knowledge from over 50 sources while preserving data sovereignty through self-hosted deployments. Unlike traditional methods, Onyx offers a more reliable, low-latency, and cost-effective context for queries, whether simple or complex research tasks.

By fetching knowledge across its internal representation, Onyx filters down to the most relevant documents, eliminating the need for iterative searching and token consumption like other agents. Onyx supports all major LLM providers, both self-hosted and proprietary, and ensures secure deployment with air-gapped, self-contained services for document index, database, and processing.

Compatible with Docker, Kubernetes, Helm/Terraform, and various cloud providers, Onyx offers two deployment options: standard and lite. The standard version includes the full feature set suitable for serious users and larger teams, while the lite mode serves as a lightweight AI chat UI requiring less memory and a simpler stack. For those who want to test Onyx quickly or have teams interested in Chat UI and Agents functionalities, Onyx Cloud offers a free deployment option.

Onyx caters to teams of all sizes, from individual users to large enterprises, and welcomes contributions via its Contribution Guide.