Software
Self-Hosted Open WebUI for a 3-Person Team
We ran Open WebUI as an internal chat and RAG workspace for a week. Here's the setup, cost, and friction.
Open WebUIself-hostingRAGteam workspaceDocker
Self-Hosted Open WebUI for a 3-Person Team
Hardware: Dell OptiPlex 7090, RTX 3060 12 GB, 64 GB RAM Software: Docker, Ollama, Open WebUI, local document directory mounted for RAG Duration: 7 days
What worked
- Single-container deployment took under 10 minutes.
- Local document RAG was accurate for internal PDFs and markdown notes.
- Per-user conversation history and model selection were intuitive.
What didn't
- Concurrent heavy users noticeably slowed token generation on a single GPU.
- No built-in SSO; local account management is fine for tiny teams only.
Cost snapshot
- Hardware: already owned
- Electricity: ~$0.80/day under load
- API spend: $0 (all local)
Verdict
Open WebUI is a strong internal ChatGPT replacement for small teams with modest concurrency. Plan to add a second GPU or move heavy models to the cloud if the team grows.