Offline Chat

Offline chat

Ragmir Chat is optional. It generates cited answers from Ragmir retrieval context on the same workstation when passages must not be sent to a hosted model. Core, the CLI, TypeScript API, and MCP server do not require Chat.

Prepare and verify a profile

rgr chat setup --profile fast
rgr chat doctor --profile fast --verify
rgr chat "What evidence supports this decision?" --profile fast --offline

Profile	Use
lite	Smaller Qwen2.5 profile for lower-memory machines
fast	Default Gemma 4 profile for a stronger local answer
quality	Larger Gemma 4 profile, explicit opt-in

Setup downloads and verifies the selected GGUF under .ragmir/models/chat/<profile>. Normal answers use that local file and do not download or build a runtime.

For air-gapped use, prepare and verify the complete profile on a connected machine, copy the directory to the same ignored path, then rerun rgr chat doctor --verify before using --offline.

Visible answers retain citations. Raw model thought is never shown, logged, or stored.

Canonical source: docs/offline-chat-preload.md.