Open WebUI: a private AI chat for your team
Give your team a ChatGPT-style interface with their own accounts, roles and groups, document search, web search and tools, connected to local models through Ollama and to any OpenAI-compatible API. Run Open WebUI on a HyperDC server behind HTTPS, preinstalled as an app option or set up with our guide.
- Accounts, roles and groups for your team
- Local models with Ollama and any OpenAI-compatible API
- Document search, web search and image generation
- Tools through MCP and OpenAPI servers
- Stack
- Python 3.11/3.12, Docker image
- Default portOpen WebUI does not terminate TLS itself: put it behind a reverse proxy with HTTPS.
- 8080 (3000 in the Docker examples)
- MinimumThe interface itself is light; local models need their own memory or a GPU.
- No official figures
- Your data
- SQLite or PostgreSQL in /app/backend/data
- Official docs
- docs.openwebui.com
Connects to
- Ollama
- OpenAI-compatible APIs
- vLLM
- llama.cpp
- LocalAI
- ComfyUI
- SearXNG
- MCP
Facts from the project’s official website, documentation and repository, checked in October 2026.
Plans are being prepared
We are preparing ready-to-use plans for Open WebUI. Tell us how you will use it and how many users you expect, and we will reply with a server that fits. You can also start today on a Linux VPS and install it with our guide.
Which server size fits?
Starting points for vCPU, memory and disk. Grow the server when your data and users grow.
| Feature |
Hosted models
Open WebUI with API providers
|
Recommended With a small local model
Open WebUI and Ollama together
|
Team on a GPU
Larger models for many users
|
|---|---|---|---|
| vCPUVirtual processor cores of the server. | 2 | 4–8 | Sized to your models |
| MemoryMemory for the app, its database and the operating system. | 2–4 GB | 16 GB | Sized to your models |
| DiskImage, chats, uploaded documents and local models. | 20 GB | 100 GB | 200 GB+ |
| Users | A team | A small team | Many users |
| Server type | Linux VPS | VDS or dedicated | GPU server (ask us) |
-
Hosted models
Open WebUI with API providers
- vCPUVirtual processor cores of the server.
- 2
- MemoryMemory for the app, its database and the operating system.
- 2–4 GB
- DiskImage, chats, uploaded documents and local models.
- 20 GB
- Users
- A team
- Server type
- Linux VPS
-
Recommended
With a small local model
Open WebUI and Ollama together
- vCPUVirtual processor cores of the server.
- 4–8
- MemoryMemory for the app, its database and the operating system.
- 16 GB
- DiskImage, chats, uploaded documents and local models.
- 100 GB
- Users
- A small team
- Server type
- VDS or dedicated
-
Team on a GPU
Larger models for many users
- vCPUVirtual processor cores of the server.
- Sized to your models
- MemoryMemory for the app, its database and the operating system.
- Sized to your models
- DiskImage, chats, uploaded documents and local models.
- 200 GB+
- Users
- Many users
- Server type
- GPU server (ask us)
Open WebUI publishes no hardware figures; its main Docker image is about 1.7 GB. Memory for local models comes on top: see the Ollama page for model sizes. Use PostgreSQL when you run more than one instance.
Everything your team expects from an AI chat
Familiar for users, controllable for admins.
Accounts, roles and groups
The first account becomes the admin; new sign-ups wait for approval, and roles and groups decide who sees which models.
Chat with documents
Upload files or build knowledge collections with hybrid search and reranking, and get answers with citations.
Web search and images
Search the web with providers such as SearXNG, and generate images through ComfyUI or an image API.
Tools and MCP
Connect MCP and OpenAPI tool servers so models can call your systems.
Your data stays yours
Prompts, files and databases stay on a server you control, in the location you choose, instead of a shared SaaS account.
Full root access
Install what the app needs, change any setting and run more services next to it. Nothing is locked behind a panel.
App option or step-by-step guide
Order the server with the app installed as an option, or set it up yourself on a clean Linux server with our guide.
Near your users
Choose a data center in the United States, Europe or Asia. The order form estimates the latency from where you are to each location.
From order to first login
Order the app preinstalled on your server, or install it yourself with our guide.
-
Pick the server
Choose a size from the table above and the data center closest to the people who will use the app.
-
Add Open WebUI
Select Open WebUI as an app option when you order, or install it on a clean Ubuntu or Debian server with our guide.
-
Point a domain and enable HTTPS
Create a DNS record such as app.example.com for the server and put a reverse proxy with a free Let’s Encrypt certificate in front of the app.
-
Create the admin and add models
Sign up first to become the admin, then add Ollama or an OpenAI-compatible connection under Admin Settings.
Step-by-step setup guides
Install, secure and update the app with our guides, written for current Ubuntu and Debian releases.
-
How to run Open WebUI with Ollama using Docker Compose and HTTPS
Deploy Open WebUI with Docker Compose next to Ollama, connect them without opening port 11434, create the admin account privately, publish it over HTTPS with Caddy and keep it backed up and updated.
35 min Intermediate
Related solutions
Frequently asked questions
Who becomes the admin?
The first account created on a new instance becomes the admin. Create it right after installing; afterwards new sign-ups get the pending role until an admin approves them.
Can I run Open WebUI and Ollama together?
Yes. The :ollama image bundles both in one container, or you run them as two containers on the same server. For larger models, run Ollama on a GPU server and connect Open WebUI to it over a private connection.
How do I connect OpenAI or another API?
In Admin Settings, open Connections and add an OpenAI-compatible connection with its base URL and API key: OpenAI itself, vLLM, llama.cpp’s server, LocalAI or another provider. Environment variables only fill these settings on the first start.
How do I secure Open WebUI?
Put it behind a reverse proxy with HTTPS, because Open WebUI does not terminate TLS itself; set a fixed WEBUI_SECRET_KEY; limit CORS to your own domain; and keep sign-ups pending for approval. The project advises against exposing it directly to the internet without such protection.
SQLite or PostgreSQL?
SQLite in the data volume is the default and works well for one instance. Use PostgreSQL through DATABASE_URL when you run several instances or want separate database backups.
Does Open WebUI need a GPU?
No. The interface runs on any server; only the models need power. With hosted APIs a small VPS is enough. The :cuda image uses an NVIDIA GPU for local features such as embeddings and speech when one is available.
Give your team its own AI chat
Tell us the app, how many people will use it and where they are, and we will suggest a server for it.