# Generated by API Evangelist (build-phrasing.py). Our phrasing, not observed demand.
overlay: 1.0.0
info:
  title: API Evangelist conversational phrasing for Thehiveryiq Com Arb Compute API
  version: 1.0.0
extends: openapi/thehiveryiq-com-arb-compute-api-openapi.yml
actions:
- target: $.info
  update:
    x-apievangelist-phrasing:
      method: generated
      generated: '2026-09-26'
      generator: build-phrasing.py
      label: Generated by API Evangelist
      operations: 3
- target: $.paths['/v1/arb/compute/route'].post
  update:
    x-apievangelist-phrasing:
      intent: Route an LLM call to the cheapest fit model
      effect: write
      questions:
      - How do I send a prompt to the cheapest model that still meets my quality bar?
      - Can I force a quality tier instead of letting the shape decide?
      instructions:
      - text: Route prompt {prompt} for shape {shape}.
        slots:
          prompt: requestBody.prompt
          shape: requestBody.shape
      - text: Route {prompt} as a {shape} call at tier {quality_tier}.
        slots:
          prompt: requestBody.prompt
          shape: requestBody.shape
          quality_tier: requestBody.quality_tier
      method: generated
      generated: '2026-09-26'
- target: $.paths['/v1/arb/compute/savings'].get
  update:
    x-apievangelist-phrasing:
      intent: See cumulative model routing savings
      effect: read
      questions:
      - How much money has cost-based model routing saved so far?
      - How many routed calls failed because every candidate model was exhausted?
      instructions:
      - text: Show routing savings.
      - text: Get total spend and savings from routed calls.
      method: generated
      generated: '2026-09-26'
- target: $.paths['/v1/arb/compute/catalog'].get
  update:
    x-apievangelist-phrasing:
      intent: List routable models and their costs
      effect: read
      questions:
      - Which models can calls be routed to, and what does each cost per 1k tokens?
      - What quality tier is each model in?
      instructions:
      - text: List the model catalog with costs.
      - text: Show models grouped by quality tier, cheapest first.
      method: generated
      generated: '2026-09-26'