# Generated by API Evangelist (build-phrasing.py). Our phrasing, not observed demand.
overlay: 1.0.0
info:
  title: API Evangelist conversational phrasing for fal Model Queue API
  version: 1.0.0
extends: openapi/fal-ai-queue-api-openapi.yml
actions:
- target: $.info
  update:
    x-apievangelist-phrasing:
      method: generated
      generated: '2026-10-01'
      generator: build-phrasing.py
      label: Generated by API Evangelist
      operations: 4
- target: $.paths['/{model_owner}/{model_name}'].post
  update:
    x-apievangelist-phrasing:
      intent: Queue a generation job on a hosted model
      effect: write
      questions:
      - How do I send a prompt to a fal model like flux/schnell and get a request ID back?
      - Can I get a webhook callback when my queued image or video generation job finishes?
      - Is there a way to pass a seed so a queued generation is reproducible?
      instructions:
      - text: Queue a job on {model_owner}/{model_name} with the prompt {prompt}.
        slots:
          model_owner: path.model_owner
          model_name: path.model_name
          prompt: requestBody.prompt
      - text: Submit {prompt} to {model_owner}/{model_name} using reference image {image_url} and seed {seed}.
        slots:
          prompt: requestBody.prompt
          model_owner: path.model_owner
          model_name: path.model_name
          image_url: requestBody.image_url
          seed: requestBody.seed
      - text: Queue {prompt} on {model_owner}/{model_name} and POST the outcome to {fal_webhook} when it completes.
        slots:
          prompt: requestBody.prompt
          model_owner: path.model_owner
          model_name: path.model_name
          fal_webhook: query.fal_webhook
      method: generated
      generated: '2026-10-01'
- target: $.paths['/{model_owner}/{model_name}/requests/{request_id}/status'].get
  update:
    x-apievangelist-phrasing:
      intent: Check the status of a queued inference job
      effect: read
      questions:
      - Is my queued fal job still IN_QUEUE, IN_PROGRESS, or already COMPLETED?
      - Can I see the model's log lines while my request is still running?
      - How do I poll whether a generation request failed or was canceled?
      instructions:
      - text: Check the queue status of request {request_id} on {model_owner}/{model_name}.
        slots:
          request_id: path.request_id
          model_owner: path.model_owner
          model_name: path.model_name
      - text: Show the status of {request_id} on {model_owner}/{model_name} with running logs set to {logs}.
        slots:
          request_id: path.request_id
          model_owner: path.model_owner
          model_name: path.model_name
          logs: query.logs
      method: generated
      generated: '2026-10-01'
- target: $.paths['/{model_owner}/{model_name}/requests/{request_id}'].get
  update:
    x-apievangelist-phrasing:
      intent: Fetch the output of a completed inference job
      effect: read
      questions:
      - How do I download the generated image or video once my queued request has completed?
      - What comes back if I ask for a job's final output while it's still running?
      - Where do I get the finished output payload for a request ID I submitted earlier?
      instructions:
      - text: Get the final result of request {request_id} from {model_owner}/{model_name}.
        slots:
          request_id: path.request_id
          model_owner: path.model_owner
          model_name: path.model_name
      - text: Fetch the generated output for job {request_id} once it finishes.
        slots:
          request_id: path.request_id
      method: generated
      generated: '2026-10-01'
- target: $.paths['/{model_owner}/{model_name}/requests/{request_id}/cancel'].put
  update:
    x-apievangelist-phrasing:
      intent: Cancel a queued or running inference job
      effect: destructive
      questions:
      - Can I stop a generation job that's still waiting in the queue or already running?
      - Does canceling an in-progress fal request free up the GPU it was using?
      instructions:
      - text: Cancel request {request_id} on {model_owner}/{model_name}.
        slots:
          request_id: path.request_id
          model_owner: path.model_owner
          model_name: path.model_name
      - text: Abort the in-progress job {request_id} so it stops using GPU capacity.
        slots:
          request_id: path.request_id
      method: generated
      generated: '2026-10-01'