Plans

Run Ragen on your own infrastructure

You install Ragen on your own infrastructure, and documents and conversations never leave it. Start on your own and for free, then reach for an SLA and rollout support only once you need them.

Self-hosted

For teams that want to run a company AI assistant on their own servers and manage it themselves.

Free

No licence fees. You only pay for your own infrastructure and the models you use.

What you get:

  • Docker image you install on your own infrastructure
  • AI assistant answering from your documents (RAG)
  • Specialized AI assistants per department
  • Website chatbot
  • MCP integrations – Drive, Gmail, HubSpot, ClickUp, Analytics and the rest
  • Any model LiteLLM supports – locally on your own GPU, or through your own keys to commercial providers
  • Roles and permissions per department
  • Deployment documentation and community support
Start with the docs or book a demo →
⭐ With support

Enterprise

For organisations that need auditing, an SLA and hands-on rollout support, including in regulated industries.

Contact us

Priced individually, based on scale and compliance requirements.

Everything in Self-hosted, plus:

  • Hands-on on-premise rollout – your own servers or your AWS, Azure or GCP account
  • Custom MCP – integrations built to order for your internal systems
  • White-label chatbots – your own domain and branding
  • Automatic PII masking – national IDs, tax numbers, card numbers
  • BYOK / CMEK – encryption keys managed in your own KMS (AWS / Azure / GCP)
  • BYO-LLM – connect your own Azure OpenAI or AWS Bedrock account
  • Audit logs and per-team spending limits
  • SLA and priority support – dedicated Slack channel
  • Dedicated Customer Success Manager
  • Annual security audits
  • Consultant-led onboarding: AI Mapping workshop, document ingest, assistant configuration
Talk to our team or book a demo →

Why this works differently from Copilot or Glean

Microsoft Copilot is $30 per employee per month. Glean starts at $40. For a 50-person company that's $1,500–2,000 monthly just for licenses, before any actual deployment work. And all of it on somebody else's infrastructure.

Ragen runs on yours. You do not pay per seat, your documents never leave your network, and when you switch model provider you switch the provider – not the whole system.

In every variant

What you get either way

🔒 Full control over your data – it stays on your infrastructure
🇪🇺 Ragen understands questions in your language even when documents use different phrasing
🔐 Encryption of conversation content
📊 Real-time AI monitoring
🔄 Zero vendor lock-in
✅ Per-file and per-folder permissions
📞 Onboarding in your language

FAQ

Common questions about the plans

Is Ragen really free?
The self-hosted variant carries no licence fees, you run Ragen from a Docker image on your own infrastructure, with no time limit. You cover the cost of your servers and of the models you use. You pay us only when you need Enterprise: audit logs, an SLA and rollout support.
What do I need to run Ragen myself?
A server with Docker – your own, or one in your AWS, Azure or GCP account – plus access to models: locally through Ollama, or via your own OpenAI, Anthropic or Azure OpenAI keys. Requirements and step-by-step setup are covered in the documentation.
Do my documents leave my network?
No. Ragen runs entirely on your infrastructure – documents, the vector store and conversation history all stay inside your network. The only outbound traffic is model requests, and only if you deliberately choose an external provider instead of a locally hosted model.
How is Enterprise different from self-hosted?
The engine is the same. Enterprise adds what security and compliance teams ask for: audit logs, PII masking, encryption keys held in your own KMS (BYOK/CMEK), an SLA with priority support and a dedicated Customer Success Manager. Plus hands-on help with the rollout itself and MCP integrations built to order.
Will you help with the rollout?
Yes, as part of Enterprise. We start with an AI Mapping workshop to identify which processes to automate. Then we ingest and optimize your documents, configure AI assistants for your specific use cases, and wire up MCP integrations.
How does Enterprise billing work?
Pricing is individual – it depends on the scale of the rollout and your compliance requirements. We invoice with 30- or 60-day payment terms. We agree the scope in a 30-minute call before you sign anything.

Not sure which variant fits your company?

In a 30-minute demo we'll work out which documents to load first, which integrations to connect at the start, and whether your team can handle the rollout alone or is better off doing it with us.

Book a demo →