openapi: 3.1.0
info:
  title: Pika Labs via fal.ai Queue Cancel Submission API
  description: 'Pika Labs does not publish a first-party REST API on its main domain.

    Production REST access to Pika''s video generation models is provided

    through the fal.ai inference platform''s standard Queue API.


    Submit an inference request with POST /{model_id}, then poll

    GET /{model_id}/requests/{request_id}/status (or stream it) until the

    request completes, and fetch the result with

    GET /{model_id}/requests/{request_id}. Authentication is a fal.ai API

    key sent as `Authorization: Key $FAL_KEY`.


    The {model_id} path component identifies a specific Pika model on fal.ai

    (for example `fal-ai/pika`). Only the generic Queue API surface is

    modeled here; see the fal.ai documentation under externalDocs for

    model-specific input and output schemas.

    '
  version: 1.0.0
  contact:
    name: fal.ai documentation
    url: https://fal.ai/docs/model-endpoints/queue
  license:
    name: fal.ai Proprietary
servers:
- url: https://queue.fal.run
  description: fal.ai inference queue base URL.
security:
- falKey: []
tags:
- name: Submission
  description: Submit inference requests.
paths:
  /{model_id}:
    parameters:
    - in: path
      name: model_id
      required: true
      schema:
        type: string
      description: 'fal.ai model identifier (e.g. `fal-ai/pika`).

        '
    post:
      tags:
      - Submission
      summary: Submit an inference request
      description: 'Submits a generation request for the specified Pika model on fal.ai.

        Returns a request identifier along with URLs to check status, fetch

        the result, and cancel the request.

        '
      operationId: submitRequest
      parameters:
      - in: query
        name: fal_webhook
        schema:
          type: string
          format: uri
        description: Optional webhook URL invoked when the request completes.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
              description: Model-specific input payload.
              additionalProperties: true
      responses:
        '200':
          description: Request accepted into the queue.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/QueueSubmission'
components:
  schemas:
    QueueSubmission:
      type: object
      properties:
        request_id:
          type: string
        response_url:
          type: string
          format: uri
        status_url:
          type: string
          format: uri
        cancel_url:
          type: string
          format: uri
        queue_position:
          type: integer
  securitySchemes:
    falKey:
      type: apiKey
      in: header
      name: Authorization
      description: 'fal.ai API key sent as `Authorization: Key $FAL_KEY`.

        '
externalDocs:
  description: fal.ai Queue API documentation
  url: https://fal.ai/docs/model-endpoints/queue