The tension
The exits, in the order the objections are raised
Running a model needs the local Pharos server.This copy of the explorer is a frozen bundle with no backend, so it cannot call a model. Install Pharos and start the explorer locally:
uv sync --group ui uv run python -m pharos.cli serve
Then open http://127.0.0.1:8080. The model view lists the models you have pulled with Ollama.