Local AI · Earlier experiment · April 2026
Trying a local chat interface with Open WebUI
Open WebUI was one of the interfaces I tried around Ollama. The setup notes are also a useful reminder that an installed app and a complete workflow are different milestones.
Before settling on a simpler LM Studio and AnythingLLM setup, I explored Ollama with Open WebUI on my Mac. An April checkpoint records Docker, Ollama and Open WebUI installed and running. It also records that I hadn’t downloaded models at that stage.
A separate set of notes describes the setup using DuckDuckGo for web search and looks at replacing it with SearXNG. Search had become another part of the experiment: how would the local chat interface find information beyond the model itself?
I kept the proposed search change as a guide while I worked through how the pieces should fit together. Later, I chose to reduce the overlapping tools I maintained and concentrate on LM Studio and AnythingLLM. Open WebUI belongs in that earlier part of the journey.
From my project notes
What came out of it
Ollama and Open WebUI were installed and running at the recorded April checkpoint; I also documented a possible change to the search setup.