Onyx
A self-hosted enterprise search and knowledge platform that connects company data to chat, agents, actions, and MCP clients.
By Onyx · reviewed 2026-10-21
What Onyx does
Onyx, formerly Danswer, indexes knowledge from company applications while preserving document metadata and access controls. Its Community Edition provides RAG, chat, agents, actions, and a broad connector catalog; it can also expose knowledge through MCP. The standard deployment is more operationally involved than a personal document-chat application.
Pricing model
Free Community Edition with paid cloud and enterprise features
Position in the stack
Where Onyx fits
Onyx is a team knowledge and context platform. It connects business systems, indexes content and permissions, retrieves evidence for chat and agents, and exposes context to external clients through MCP.
Typical deployment flow
A practical sequence for evaluating Onyx before making it part of a permanent stack.
- STEP 1
Define the data boundary
Select repositories and verify identity, permission, retention, and deletion requirements.
- STEP 2
Deploy indexing services
Choose Lite for interface evaluation or Standard for connectors and retrieval.
- STEP 3
Evaluate retrieval
Test filtering, freshness, citations, parsing, reranking, and failures on real questions.
Best for
- Team knowledge search across many business systems
- Permission-aware enterprise RAG
- Providing company context to agents through MCP
Not the right layer for
- A lightweight personal document chat
- Teams without capacity to operate indexing and storage services
Capabilities
Capabilities refer to the tool's application layer. Hardware and model support can still depend on a connected inference engine.
Deployment and compatibility
- Deployment
- Self-hostedManaged cloudCluster
- Platforms
- Web · Linux · Docker · Kubernetes
- Hardware backends
- CPU · NVIDIA GPU · Uses connected model provider
- Install methods
- Install script · Docker Compose · Kubernetes · Managed cloud
Models and integrations
- Model formats
- Uses connected local or hosted model provider · Embedding models · Reranking models
- Common integrations
- Google DriveSlackConfluenceGitHubJiraMCPOpenAI-compatible APIs
Strengths
- Large connector catalog
- Permission-aware retrieval and enterprise controls
- Knowledge access for both people and agents
Limitations
- Standard deployment has several infrastructure components
- Some enterprise governance features are outside Community Edition
Planning checklist
Before you choose Onyx
Answer these questions with the exact models, hardware, users, and data you expect to operate.
- Which source permissions must be preserved?
- How quickly must changes propagate?
- Which embedding and reranking models will be used?
- Can the team operate the services and backups?
Onyx FAQ
What is Onyx best at?
Connecting team data sources into permission-aware search, chat, agents, and MCP context.
Is Community Edition open source?
Yes. Core Community Edition code is MIT licensed; enterprise code has separate terms.
Is it lightweight?
Lite mode is small, but full indexing requires a multi-service deployment.
Official sources
Use these links to confirm current compatibility and installation requirements.
Related tools
A desktop and self-hosted workspace for document chat, RAG, and AI agents.
A self-hostable personal AI and knowledge assistant with document search, agents, automations, and research workflows.
A self-hosted multi-provider chat application with agents, files, search, MCP, and model routing.
Comparisons featuring Onyx
Use a reviewed comparison when the choice is between two adjacent tools.
Execution evidence
Known working recipes using Onyx
Recipes connect hardware, a model artifact, tools, settings, verification, and a reportable result.
No verified recipe is linked to this record yet.
Compatibility estimates remain available in the planner. A recipe appears here only after its exact stack and verification protocol are documented.