terraform-provider-stackgen

sg_guild_model (Data Source)

⚠️ Experimental — Looks up an existing Guild model by name. Use this to reference pre-provisioned LLM model registrations from agents or workflows without managing sg_guild_model in the same stack. When provider workspace_id is set, the lookup is scoped to that workspace.

Example Usage

data "sg_guild_model" "toolcalling" {
  name = "gpt-4o-toolcalling"
}

output "guild_model_provider_name" {
  value = data.sg_guild_model.toolcalling.provider_name
}

Schema

Required

• name (String) Unique model name to look up (e.g. gpt-4o-toolcalling or openai-gpt-5.4-mini).

Read-Only

• created_at (String) Creation timestamp for this registered model (RFC 3339).
• max_tokens (Number) Maximum response tokens configured for the model when returned by the API.
• model_id (String) Provider-specific model identifier (e.g. gpt-4o or claude-sonnet-4-6).
• pricing_input_usd (Number) Input-side price from the registered model when returned by the API.
• pricing_output_usd (Number) Output-side price from the registered model when returned by the API.
• pricing_total_usd (Number) Total price hint from the registered model when returned by the API.
• provider_name (String) Name of the registered model provider this model belongs to.
• reasoning_effort (String) Vendor-specific reasoning effort configured for this model; null when unset.
• scope (String) Integration scope for this model: TENANT PROJECT or USER.
• temperature (Number) Sampling temperature configured for the model when returned by the API.
• thinking (Boolean) Whether Guild routes Think/plan/synthesis work to this model; false means NoThink/dig/classify routing.
• updated_at (String) Last update timestamp for this registered model (RFC 3339).
• variant (String) Provider-specific variant or version tag returned by the API.