Airtop Page Interaction API

openapi: 3.0.1
info:
  title: Airtop AI Query Page Interaction API
  description: REST API for Airtop, a cloud-browser platform for AI agents. Create and manage remote Chromium browser sessions, open and navigate windows, drive pages with natural-language interactions (click, type, hover, scroll), query and extract page content with AI, capture screenshots, and persist state with profiles. All requests are authenticated with a Bearer API key.
  termsOfService: https://www.airtop.ai/legal/terms-of-service
  contact:
    name: Airtop Support
    url: https://docs.airtop.ai
  version: '1.0'
servers:
- url: https://api.airtop.ai/api/v1
  description: Airtop production API
security:
- bearerAuth: []
tags:
- name: Page Interaction
  description: Drive a page with natural-language click, type, hover, and scroll.
paths:
  /sessions/{sessionId}/windows/{windowId}/click:
    post:
      operationId: click
      tags:
      - Page Interaction
      summary: Click
      description: Clicks an element identified by a natural-language description. AI resolves the description to an on-page target before clicking.
      parameters:
      - $ref: '#/components/parameters/SessionId'
      - $ref: '#/components/parameters/WindowId'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ClickRequest'
      responses:
        '200':
          description: Click performed.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/AIResponse'
  /sessions/{sessionId}/windows/{windowId}/hover:
    post:
      operationId: hover
      tags:
      - Page Interaction
      summary: Hover
      description: Hovers over an element described in natural language.
      parameters:
      - $ref: '#/components/parameters/SessionId'
      - $ref: '#/components/parameters/WindowId'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ElementInteractionRequest'
      responses:
        '200':
          description: Hover performed.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/AIResponse'
  /sessions/{sessionId}/windows/{windowId}/type:
    post:
      operationId: type
      tags:
      - Page Interaction
      summary: Type
      description: Types text into a field described in natural language, with options to clear the field first and press Enter afterward.
      parameters:
      - $ref: '#/components/parameters/SessionId'
      - $ref: '#/components/parameters/WindowId'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TypeRequest'
      responses:
        '200':
          description: Text typed.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/AIResponse'
  /sessions/{sessionId}/windows/{windowId}/scroll:
    post:
      operationId: scroll
      tags:
      - Page Interaction
      summary: Scroll
      description: Scrolls the page, optionally toward a described element or by an offset.
      parameters:
      - $ref: '#/components/parameters/SessionId'
      - $ref: '#/components/parameters/WindowId'
      requestBody:
        required: false
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ScrollRequest'
      responses:
        '200':
          description: Scroll performed.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/AIResponse'
components:
  schemas:
    AIResponse:
      allOf:
      - $ref: '#/components/schemas/Envelope'
      - type: object
        properties:
          data:
            type: object
            properties:
              modelResponse:
                type: string
                description: The AI model's textual response to the prompt.
          meta:
            type: object
            properties:
              requestId:
                type: string
              status:
                type: string
              usage:
                type: object
                properties:
                  credits:
                    type: number
              screenshots:
                type: array
                items:
                  type: object
    ElementInteractionRequest:
      type: object
      properties:
        elementDescription:
          type: string
          description: Natural-language description of the target element.
        clientRequestId:
          type: string
        costThresholdCredits:
          type: integer
        timeThresholdSeconds:
          type: integer
        configuration:
          type: object
      required:
      - elementDescription
    TypeRequest:
      type: object
      properties:
        text:
          type: string
          description: The text to type.
        elementDescription:
          type: string
          description: Natural-language description of the field to type into.
        pressEnterKey:
          type: boolean
          description: Press Enter after typing.
        clearInputField:
          type: boolean
          description: Clear the field before typing.
        clientRequestId:
          type: string
        costThresholdCredits:
          type: integer
        timeThresholdSeconds:
          type: integer
        configuration:
          type: object
      required:
      - text
    ScrollRequest:
      type: object
      properties:
        scrollToElement:
          type: string
          description: Natural-language description of an element to scroll to.
        scrollBy:
          type: object
          description: Pixel offset to scroll by.
        configuration:
          type: object
    Envelope:
      type: object
      properties:
        meta:
          type: object
          properties:
            requestId:
              type: string
            status:
              type: string
        errors:
          type: array
          items:
            type: object
        warnings:
          type: array
          items:
            type: object
    ClickRequest:
      allOf:
      - $ref: '#/components/schemas/ElementInteractionRequest'
      - type: object
        properties:
          configuration:
            type: object
            properties:
              clickType:
                type: string
                enum:
                - click
                - doubleClick
                - rightClick
              visualAnalysis:
                type: object
              waitForNavigationConfig:
                type: object
          waitForNavigation:
            type: boolean
  parameters:
    SessionId:
      name: sessionId
      in: path
      required: true
      description: Identifier of the cloud browser session.
      schema:
        type: string
    WindowId:
      name: windowId
      in: path
      required: true
      description: Identifier of the window within the session.
      schema:
        type: string
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      description: 'Airtop API key supplied as a Bearer token in the Authorization header: `Authorization: Bearer YOUR_API_KEY`.'