// Provider-wire compat for "firepass"; regenerated from the frozen census using deployment contract selectors.

provider "firepass" {
	default-model "glm-5.2-fast"
	env "FIREPASS_API_KEY"

	// Pricing and limits are the published Fire Pass router tariff
	// (https://docs.fireworks.ai/firepass), not upstream-discoverable: dedicated
	// `fpk_…` keys never authorize `/v1/models`, so this seed is the authoritative
	// source. cacheRead is 0.1x input; cacheWrite stays 0 (not billed separately).
	seed api="openai-completions" base-url="https://api.fireworks.ai/inference/v1" bundle="fallback" {
		model "glm-5.2-fast" name="GLM 5.2 Fast (Fire Pass)" {
			reasoning #true
			input "text"
			cost input=2.1 output=6.6 cache-read=0.21 cache-write=0
			limits context=1048576 max-tokens=131072
		}
		model "kimi-k3-fast" name="Kimi K3 Fast (Fire Pass)" {
			reasoning #true
			input "text" "image"
			cost input=4.5 output=22.5 cache-read=0.45 cache-write=0
			limits context=1048576 max-tokens=131072
		}
	}

	// Replaces the Fireworks-hosted minimal-effort remap for Fire Pass.
	thinking-effort-map {
		minimal "none"
	}
	// Replaces the Fire Pass wire-model-id dispatch baseline.
	wire-model-id-mode "firepass"
	class "glm" priority=1 {
		thinking-mode "effort"
		thinking-efforts "minimal" "low" "medium" "high" "max"
	}

	class "kimi" priority=1 {
		thinking-mode "effort"
		thinking-efforts "low" "high" "max"
	}
	// The Fire Pass router shares Fireworks' 32,768 Kimi K2 output cap (#1849).
	class "kimi" {
		family "k2" {
			max-tokens-ceiling 32768
		}
		family "k2.5" {
			max-tokens-ceiling 32768
		}
		family "k2.6" {
			max-tokens-ceiling 32768
		}
		family "k2-thinking" {
			max-tokens-ceiling 32768
		}
	}
}