Models and routing

Providers answer where Gormes sends model requests. Models and routing answer which model is selected for a run and how overrides behave.

Choose interactively

gormes setup model
gormes model

Override for one run

gormes chat --provider openai --model gpt-5.5
gormes chat -q "hello" --provider openai --model gpt-5.5

Invocation flags override environment variables, which override config.toml, which override built-in defaults.

Fallback chains

Use fallback when you want resilience across providers:

gormes fallback --help

The operator workflow lives in Fallback provider chain. Exact provider credential setup lives in Providers.