The Butler practical library
Fix the problem in front of you.
Short, specific diagnosis paths for a local AI setup that is not behaving.
Reviewed November 14, 2026. Official-documentation-based guidance; not a claim that every installation or hardware combination was hands-on tested.
3 fixes available
Ollama connection refused
Check whether the local server is running before changing networking.
Explore →The model will not fit in memory
Separate RAM, graphics memory and context before downloading another model.
Explore →Open WebUI cannot reach Ollama
The browser, container and model server may not share the same localhost.
Explore →No matching fix yet. Try another word or check your tool’s setup guide.
Do not post API keys, session files or private document contents when asking for help. Include the exact error, app version, operating system and whether Docker is involved.