RTX 5060 Ti 16GB OpenWebUI Setup

OpenWebUI gives you a polished ChatGPT-style frontend to your self-hosted LLM. Connect it to vLLM or Ollama on the RTX 5060 Ti 16GB at our hosting:

Contents

• Install
• Connect to backend
• RAG / documents
• Multi-user

Install

docker run -d \
  --name openwebui \
  -p 3000:8080 \
  -e WEBUI_NAME="My AI" \
  -v openwebui:/app/backend/data \
  --restart unless-stopped \
  ghcr.io/open-webui/open-webui:main

Access at http://server:3000. First user to register becomes admin.

Connect to Backend

In OpenWebUI Admin -> Settings -> Connections:

• vLLM: OpenAI API endpoint = http://host.docker.internal:8000/v1 (or container IP)
• Ollama: Ollama API URL = http://host.docker.internal:11434

Models auto-populate in the chat dropdown once connected.

RAG / Documents

OpenWebUI has built-in document ingestion:

• Drop PDFs / text files – chunks and embeds automatically
• Use # to tag documents in chat for retrieval-augmented answers
• Configure embedding model in Admin -> Settings (default SentenceTransformers)
• Point to your self-hosted TEI embedding server for consistent quality

Multi-User

• Invite team via admin panel or enable self-signup
• Per-user chat history stored in OpenWebUI’s SQLite
• OAuth / SAML integration for SSO
• Roles: admin, user, pending – gate access to models/features

For a team of 5-20 running a self-hosted ChatGPT replacement on one 5060 Ti, OpenWebUI is the most polished frontend available.

See also: vLLM setup, Ollama setup, embedding server, internal tooling.