Discover local models in GitHub Copilot CLI
By Steven Van ·
CLI version 1.0.94-0 lets users add and switch to an installed Ollama model during a session without restarting.
Local Ollama models now appear alongside configured models and Copilot cloud models in the /model picker in GitHub Copilot CLI, starting with version 1.0.94-0.
Add a model during a session
Discovery does not automatically add models. Select a discovered model, review its provider and endpoint, then choose “Add and use for this session” or “Add without switching”. Switching to the model does not require restarting the CLI. If a provider connection fails, the picker displays an explanation.
Local model requirements
Ollama and the model must already be installed and running. The picker does not install a runtime or download models. Models must support tool calling and streaming; the documentation recommends a context window of at least 128k tokens for best results.
Offline mode remains a separate setting
Choosing a local model does not enable offline mode or disable GitHub telemetry. Offline mode requires COPILOT_OFFLINE=true to prevent the CLI from contacting GitHub’s servers. A remote provider can still receive prompts and code context over the network, even with offline mode enabled. The provider and offline-mode documentation covers configuration and limits.
