Model Selection & Token Modes

Supported Providers

YeePilot supports:

• OpenRouter
• OpenAI
• Anthropic

Set Default Provider/Model

Use setup:

yeepilot setup --provider

Or configure directly in ~/.yeepilot/config.yaml:

ai:
  provider: openrouter
  model: openai/gpt-oss-120b:free

Per-Command Override

yeepilot exec --provider openai --model gpt-4o "summarize this folder"

This override applies only to that one-shot run.

In-Session Model Switch

Use /model in TUI to view/switch model/provider selection.

Token Modes

normal

Balanced/default behavior.

saver

Lower token usage with more compact behavior for long sessions.

Switch in session:

/tokenmode normal
/tokenmode saver

Set default in config:

ai:
  token_mode: saver

Practical Recommendations

• Use smaller/cheaper models for routine checks.
• Use stronger models for architecture/refactor/security reviews.
• Pair expensive models with /tokenmode saver on long sessions.