Onyx AI | Open Source Enterprise Search & AI Assistant

Self-hosted AI Chat and Enterprise Search

Work with AI agents connected to your internal knowledge and tools.

Strict permission controls and your data never leaves your network.

Trusted by security-conscious teams worldwide.

Case Study

How Ramp Deployed Secure AI to Thousands of Employees

Learn how one of the fastest-growing fintechs uses Onyx to answer thousands of questions a week with enterprise-grade security.

Download the case study

Deployment

Your infrastructure. Your rules.

Onyx can run entirely inside your network boundary. Run open-source LLMs locally or connect to any model provider.

SOC 2 Type II certified. Currently deployed in FedRAMP, ITAR, CMMC, FERPA, and GDPR-regulated environments.

VPC

Your data center

Zero internet dependency

Your own hardware

Connectors

Permission-aware from day one

45+ connectors that don't just index your data, they mirror your fine-grained access controls. Every query respects who can see what.

Document-level permissions

Onyx syncs ACLs from every connected source. Users only see documents they are authorized to access.

Real-time permission sync

When access changes in SharePoint, Confluence, or Google Drive, Onyx reflects it within minutes. No stale permissions. No manual re-indexing.

SCIM and IdP integration

Provision users and groups from Okta, Azure AD, or any SAML/OIDC provider. Role-based access maps directly to your existing identity infrastructure.

Capabilities

Search, chat, and agents on your terms.

Give every member of your team a personal AI assistant for research, chat, and automations. Self-hosted or on our managed cloud.

Search
One search bar across Slack, Confluence, SharePoint, Google Drive, Jira, and dozens more. Get exact answers with citations, or run Deep Research for in-depth analysis across your entire knowledge base.

Chat
AI chat tailored to your workflow. Agents that draw on your real-time data and tools to answer questions and take action.

Platform
Embed Onyx into your own products and workflows. Full REST API, an MCP server, and a connector framework for custom data sources.

Governance

Deploy AI broadly. Govern it tightly.

Control which models your org uses. Onyx works with any LLM and provider. OpenAI, Anthropic, Azure, Bedrock, or local open-weight models. Approve specific models per team or use case. Swap providers without disrupting users.

See AI usage across your organization. Usage analytics, query logs, and feedback tracking across every team.

Admin controls and rate limits. Set per-user and per-group rate limits, restrict access to specific connectors, and enforce usage policies.

Eliminate shadow AI. Give your team an approved AI platform that actually works.

Ready to deploy secure AI?

Self-hosted or managed cloud. Your data, your infrastructure.

Book a Demo
See the Repo

Secure

Enterprise-grade security and compliance. Flexible and secure deployment options.

Open-Source

Deep customizability optimized for your use case, supported by a large Open-Source community.

Stars
31k

Docs

Trusted by top teams

Case study

Self-Hosted AI Platform for Enterprise