Skip to content
CartControl panel

Open WebUI + Ollama 

A local AI interface with models you choose.

On your VPS. Maintained by you.

Self hosted, self managed AI stack

RAG, knowledge and tools

Ollama as local model space

A self hosted, self managed local AI stack on our VPS platform

Your Open WebUI + Ollama environment on TransIP VPS: download, run and use AI models from a clear web interface. You maintain the entire stack yourself, from installation and updates to security. Build a local AI stack for chat, knowledge, self hosted RAG and tools, while staying in control of models, documents, access and infrastructure.

Select the right VPS for your Open WebUI + Ollama server

Heavier local models require more resources. More users, bigger knowledge bases and embeddings add to that. For a first Open WebUI + Ollama environment a V3 is a good starting point.

V3

  • 2 shared or dedicated vCPU's
  • 4 GB RAM
  • 100 GB NVMe
  • Unlimited traffic
  • 1 Snapshot

From

€ 20,-per month

Configure

V4

  • 4 shared or dedicated vCPU's
  • 8 GB RAM
  • 100 GB NVMe
  • Unlimited traffic
  • 1 Snapshot

From

€ 50,-per month

Configure

V5

  • 8 shared or dedicated vCPU
  • 16 GB RAM
  • 100 GB NVMe
  • Unlimited traffic
  • 1 Snapshot

From

€ 100,-per month

Configure
 

↓

Open WebUI + Ollama: a powerful combination

Features

Chat

Talk to multiple local models from one interface and compare answers in the same workflow.

Self hosted RAG

Upload documents and let Open WebUI build answers based on your own knowledge base with retrieval.

Local models

Let Ollama download models, run and serve them on your own VPS without external LLMs.

Teams & access

Work with roles, groups and shared access in a multi-user environment.

Tools

Add web search, code execution and other tools to chats and workflows on top of local models.

API & extensions

Expand with Python functions, pipelines, MCP and OpenAPI servers and connect with the Ollama API.

Open WebUI and Ollama work together immediately

Open WebUI has its own Ollama connection to maintain models from the admin omgeving. From there you can download models, connect multiple Ollama instances and use those models for chats, knowledge bases and tools in the same interface.
Kies jouw VPS

More options with knowledge and RAG

Open WebUI puts documents, collections and retrieval under your control. Upload PDFs, spreadsheets, code or other text files and let your models use RAG to retrieve relevant pieces from your own knowledge base. That way you build a self-hosted AI experience that doesn't just chat, but also answers from your own context.
Kies uit drie pakketten

Self hosted, self managed on your VPS

Open WebUI + Ollama on TransIP VPS is fully self managed: you host the models, conversations, documents and extensions, giving full control of privacy and configuration. TransIP provides the VPS, but the installation, updates, back-ups, access, monitoring and security of Open WebUI and Ollama is all up to you.
Kies uit drie pakketten

Want to know more?

In the Open WebUI documentation you can find manuals for connecting to Ollama, information on knowledge bases, RAG and maintaining your self hosted AI environment.

Zoeken in de Knowledge Base

FAQ

  • What is Open WebUI + Ollama?

    Open WebUI + Ollama is a self hosted, self managed AI stack where Open WebUI provides the interface and Ollama downloads local AI models and runs them. Together they form a local AI environment on your VPS, with full control over data and access. You maintain the entire stack yourself.

  • How does it work?

    Open WebUI runs on your VPS and connects via a local API with Ollama. Questions go from the interface to the model and answers can be enriched with your own data via RAG, making results more relevant.

  • What should you be mindful of?

    Self hosting requires sufficient resources and proper maintenance. Keep RAM and CPU requirements in mind for models, secure your server and take care of updates, back-ups and scalability as your usage grows. Also keep in mind that you maintain the entire stack yourself, there's no managed layer or application support from TransIP.