What is the minimum RAM for Open WebUI?

Open WebUI itself needs about 512 MB of RAM. If you use Ollama with a local model (7B), plan for at least 8 GB of RAM. For the OpenAI API only, 2 GB is enough.

Yes, Open WebUI is open source (MIT). You only pay for the VPS resources. If you use the OpenAI API, the API calls are billed by OpenAI — not by Open WebUI.

How does it differ from ChatGPT?

ChatGPT is hosted by OpenAI and all your data is sent to it. Open WebUI is self-hosted on your server: zero data transfer, full privacy, local models possible.

Can several models be used at once?

Yes. Open WebUI supports multiple connections (Ollama + OpenAI + Anthropic in parallel). You select the model for each conversation.

Build, host and run AI solutions.

Open WebUI

Web interface for your LLMs — Ollama, OpenAI, Mistral — hosted on your own server.

2 GB (8 GB with Ollama) RAM 2 vCPU Port 3000 Available

Source code

Tech stack

DockerPythonOllama (optional)

Minimum RAM2 GB (8 GB with Ollama)

Minimum CPU2 vCPU

Default port3000

Compatible OSubuntu-24.04

Open WebUI is an open-source chat interface, compatible with Ollama and the OpenAI APIs, that lets you interact with language models from your browser. Deployed on your VPS, it replaces ChatGPT without sending your data to third parties: your conversations, documents and history stay on your infrastructure.

It is the ideal tool for teams that want to experiment with local LLMs (Llama 3, Mistral, Gemma) or remote ones while keeping full control of their exchanges. Open WebUI supports RAG (Retrieval-Augmented Generation), multi-model conversations and session sharing between users.

Key features

Compatible with Ollama (local models) and the OpenAI/Anthropic API

Built-in RAG: query your PDF documents, texts and knowledge bases

Multi-user management with roles and permissions

Persistent conversation history, exportable

Multimodal support: text, images, code

Responsive interface, light/dark theme

When to use this solution?

1

Internal team

Deploy a shared AI assistant for your team, connected to your internal documents, without exposing your data to OpenAI.

2

Development & testing

Test different LLM models side by side before integrating one of them into your application.

3

Intelligent knowledge base

Upload your PDFs, procedures and technical docs — ask questions in natural language.

Deploy Open WebUI on your VPS

Guide optimized for ServOrbit Cloud VPS.

01

Create the VPS

Order a ServOrbit Cloud VPS (minimum 4 GB RAM if you use Ollama with a local model). Choose Ubuntu 22.04.

02

Install Docker

Update the system and install Docker + Docker Compose. Open WebUI runs entirely through containers.

03

Deploy Open WebUI

Run docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway ghcr.io/open-webui/open-webui:main. The interface is reachable on port 3000.

04

Configure Nginx + SSL

Set up an Nginx reverse proxy with Let's Encrypt to expose Open WebUI on your domain over HTTPS.

05

Connect your models

In the settings, add the URL of your Ollama instance (or your OpenAI API key). Your models appear immediately in the conversation selector.

Frequently asked questions

Open WebUI is a self-hosted web interface for interacting with LLMs (Large Language Models) such as Llama, Mistral or GPT-4. It works like ChatGPT but runs on your own server.

Complementary solutions

Artificial Intelligence

Activate Open WebUI on your infrastructure.

Dedicated Cloud VPS — IPv4 included, European datacenter, support included. Your data never leaves your server.

Recommended configuration: 2 GB (8 GB with Ollama) RAM · 2 vCPU