Ollama is disconnected
Start Ollama, keep the host reachable, then run LocalPilot: Check Ollama Status.
ollama serveTroubleshooting
Fix common LocalPilot setup and runtime issues
Most issues come down to Ollama reachability, missing local models, model size, or safety filters doing their job.
No models are found
Run LocalPilot: Run Setup or pull a model manually, then select it inside VS Code.
ollama pull qwen2.5-coder:1.5bInline suggestions feel slow
Switch inline completion mode to line, use micro or lite mode, lower context/output limits, or choose a smaller model.
localpilot.inlineCompletionMode = line
localpilot.mode = liteOutput quality is weak
Try a larger coding model and select it for chat or inline suggestions with LocalPilot: Select Local Model.
ollama pull qwen2.5-coder:7bFiles are skipped
LocalPilot intentionally blocks secrets, generated folders, dependency folders, lock files, minified files, and oversized files.
localpilot.maxFileSizeKb = 500