openapi: 3.1.0
info:
  title: Runway Video Generation API
  description: >-
    The Runway Video Generation API allows developers to generate videos from
    text prompts, images, or existing videos using Runway's Gen-4, Gen-4.5,
    Gen-4 Turbo, Aleph, and Veo models. The API supports text-to-video,
    image-to-video, video-to-video, character performance (Act Two), lip sync,
    video upscale, frame interpolation, and sound effect generation workflows.
    Tasks are processed asynchronously, with developers submitting generation
    requests and polling for results using unique task identifiers.
    Authentication is handled via Bearer token in the HTTP Authorization header.
  version: '2024-11-06'
  contact:
    name: Runway Support
    url: https://support.runwayml.com/
  termsOfService: https://runwayml.com/terms-of-use
externalDocs:
  description: Runway API Documentation
  url: https://docs.dev.runwayml.com/api/
servers:
  - url: https://api.dev.runwayml.com/v1
    description: Production Server
tags:
  - name: Character Performance
    description: >-
      Control a character's facial expressions and body movements using a
      reference performance video with the Act Two model.
  - name: Frame Interpolation
    description: >-
      Interpolate between video frames to increase frame rate and smoothness.
  - name: Image to Video
    description: >-
      Generate videos from input images with optional text prompts using Gen-4,
      Gen-4 Turbo, Gen-4.5, or Aleph models.
  - name: Lip Sync
    description: >-
      Create generative videos where a selected face speaks lines from audio
      clips or AI-generated voices, supporting 28+ languages.
  - name: Sound Effects
    description: >-
      Generate sound effects for videos using AI.
  - name: Tasks
    description: >-
      Retrieve status and output of asynchronous generation tasks, or cancel
      pending tasks.
  - name: Text to Video
    description: >-
      Generate videos from text prompts alone using Gen-4.5, Veo 3.1, or
      Veo 3.1 Fast models.
  - name: Uploads
    description: >-
      Upload temporary media files that can be referenced in generation requests.
  - name: Video to Video
    description: >-
      Generate new videos from existing video inputs using the Gen-4 Aleph model.
  - name: Video Upscale
    description: >-
      Upscale video resolution and quality.
security:
  - bearerAuth: []
paths:
  /image_to_video:
    post:
      operationId: createImageToVideo
      summary: Create image-to-video generation task
      description: >-
        Starts a new asynchronous task to generate a video from an input image.
        Optionally accepts a text prompt to guide the generation. Supports Gen-4,
        Gen-4 Turbo, Gen-4.5, and Aleph models. Returns a task ID that can be
        polled for completion.
      tags:
        - Image to Video
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ImageToVideoRequest'
      responses:
        '200':
          description: Task created successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreatedResponse'
        '400':
          description: Bad request - invalid parameters
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Rate limit exceeded
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
  /text_to_video:
    post:
      operationId: createTextToVideo
      summary: Create text-to-video generation task
      description: >-
        Starts a new asynchronous task to generate a video from a text prompt
        alone, without requiring an input image. Supported by Gen-4.5, Veo 3.1,
        and Veo 3.1 Fast models. Returns a task ID that can be polled for
        completion.
      tags:
        - Text to Video
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TextToVideoRequest'
      responses:
        '200':
          description: Task created successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreatedResponse'
        '400':
          description: Bad request - invalid parameters
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Rate limit exceeded
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
  /video_to_video:
    post:
      operationId: createVideoToVideo
      summary: Create video-to-video generation task
      description: >-
        Starts a new asynchronous task to generate a video from an existing video
        input using the Gen-4 Aleph model. The input video serves as a reference
        for the generation, with a text prompt guiding the transformation.
        Returns a task ID that can be polled for completion.
      tags:
        - Video to Video
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/VideoToVideoRequest'
      responses:
        '200':
          description: Task created successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreatedResponse'
        '400':
          description: Bad request - invalid parameters
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Rate limit exceeded
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
  /character_performance:
    post:
      operationId: createCharacterPerformance
      summary: Create character performance (Act Two) task
      description: >-
        Starts a new asynchronous task to control a character's facial
        expressions and body movements using a reference performance video.
        Uses the Act Two model to animate a character image or video based on
        a driving reference. Returns a task ID that can be polled for completion.
      tags:
        - Character Performance
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CharacterPerformanceRequest'
      responses:
        '200':
          description: Task created successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreatedResponse'
        '400':
          description: Bad request - invalid parameters
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Rate limit exceeded
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
  /lip_sync:
    post:
      operationId: createLipSync
      summary: Create lip sync generation task
      description: >-
        Starts a new asynchronous task to create a generative video where a
        selected face speaks lines from audio clips or AI-generated voices.
        Supports 28+ languages using the eleven_multilingual_v2 model. Returns
        a task ID that can be polled for completion.
      tags:
        - Lip Sync
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/LipSyncRequest'
      responses:
        '200':
          description: Task created successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreatedResponse'
        '400':
          description: Bad request - invalid parameters
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Rate limit exceeded
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
  /video_upscale:
    post:
      operationId: createVideoUpscale
      summary: Create video upscale task
      description: >-
        Starts a new asynchronous task to upscale a video to higher resolution
        and quality. Accepts a video asset and returns a task ID that can be
        polled for completion.
      tags:
        - Video Upscale
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/VideoUpscaleRequest'
      responses:
        '200':
          description: Task created successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreatedResponse'
        '400':
          description: Bad request - invalid parameters
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Rate limit exceeded
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
  /frame_interpolation:
    post:
      operationId: createFrameInterpolation
      summary: Create frame interpolation task
      description: >-
        Starts a new asynchronous task to interpolate between video frames,
        increasing the frame rate and smoothness of the output video. Returns
        a task ID that can be polled for completion.
      tags:
        - Frame Interpolation
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/FrameInterpolationRequest'
      responses:
        '200':
          description: Task created successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreatedResponse'
        '400':
          description: Bad request - invalid parameters
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Rate limit exceeded
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
  /sound_effect:
    post:
      operationId: createSoundEffect
      summary: Create sound effect generation task
      description: >-
        Starts a new asynchronous task to generate sound effects for a video
        using AI. Returns a task ID that can be polled for completion.
      tags:
        - Sound Effects
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/SoundEffectRequest'
      responses:
        '200':
          description: Task created successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreatedResponse'
        '400':
          description: Bad request - invalid parameters
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Rate limit exceeded
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
  /tasks/{id}:
    get:
      operationId: getTask
      summary: Retrieve task status and output
      description: >-
        Retrieves the current status and output of an asynchronous generation
        task by its unique identifier. Poll this endpoint until the task status
        indicates completion (SUCCEEDED or FAILED). When a task succeeds, the
        response includes output URLs linking to the generated media.
      tags:
        - Tasks
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
        - $ref: '#/components/parameters/TaskId'
      responses:
        '200':
          description: Task status retrieved successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskResponse'
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '404':
          description: Task not found
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
    delete:
      operationId: deleteTask
      summary: Cancel or delete a task
      description: >-
        Cancels a pending task or deletes a completed task by its unique
        identifier. This operation cannot be undone.
      tags:
        - Tasks
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
        - $ref: '#/components/parameters/TaskId'
      responses:
        '204':
          description: Task deleted successfully
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '404':
          description: Task not found
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
  /uploads:
    post:
      operationId: createUpload
      summary: Upload a temporary media file
      description: >-
        Uploads a temporary media file that can be referenced in subsequent API
        generation requests using a runway:// URI. The response includes an
        upload URL and form fields to complete the multipart upload. Once the
        upload completes, the runway:// URI is ready to use in generation
        requests.
      tags:
        - Uploads
      parameters:
        - $ref: '#/components/parameters/RunwayVersion'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/UploadRequest'
      responses:
        '200':
          description: Upload initiated successfully
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/UploadResponse'
        '400':
          description: Bad request - invalid filename or unsupported media type
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized - invalid or missing API key
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      description: >-
        API key passed via the HTTP Authorization header using the Bearer scheme.
        Obtain your API key from the Runway Developer Portal at
        https://dev.runwayml.com/.
  parameters:
    RunwayVersion:
      name: X-Runway-Version
      in: header
      required: true
      description: >-
        API version identifier. Must be set to the exact value 2024-11-06.
      schema:
        type: string
        enum:
          - '2024-11-06'
    TaskId:
      name: id
      in: path
      required: true
      description: >-
        The unique identifier of the task, returned when the task was created.
      schema:
        type: string
        format: uuid
  schemas:
    ImageToVideoRequest:
      type: object
      required:
        - model
        - promptImage
      properties:
        model:
          type: string
          description: >-
            The model to use for generation. Supported values include gen4,
            gen4_turbo, gen4.5, and gen4_aleph.
          enum:
            - gen4
            - gen4_turbo
            - gen4.5
            - gen4_aleph
        promptImage:
          type: string
          description: >-
            An HTTPS URL, runway:// URI, or data URI containing an encoded image
            to use as the input for video generation.
        promptText:
          type: string
          description: >-
            A text description of up to 1000 characters that describes in detail
            what should appear in the generated output video.
          maxLength: 1000
        duration:
          type: integer
          description: >-
            The duration of the generated video in seconds. Common values are 5
            and 10.
          enum:
            - 5
            - 10
        ratio:
          type: string
          description: >-
            The aspect ratio of the output video. Supported landscape ratios
            include 1280:720, 1584:672, 1104:832. Portrait ratios include
            720:1280, 832:1104. Square ratio is 960:960.
          enum:
            - '1280:720'
            - '1584:672'
            - '1104:832'
            - '720:1280'
            - '832:1104'
            - '960:960'
    TextToVideoRequest:
      type: object
      required:
        - model
        - promptText
      properties:
        model:
          type: string
          description: >-
            The model to use for text-to-video generation. Gen-4.5 supports
            text-only generation by omitting the promptImage parameter. Veo
            models also support text-to-video.
          enum:
            - gen4.5
            - veo3.1
            - veo3.1_fast
            - veo3
        promptText:
          type: string
          description: >-
            A text description of up to 1000 characters that describes in detail
            what should appear in the generated output video.
          maxLength: 1000
        duration:
          type: integer
          description: >-
            The duration of the generated video in seconds.
          enum:
            - 5
            - 10
        ratio:
          type: string
          description: >-
            The aspect ratio of the output video.
          enum:
            - '1280:720'
            - '1584:672'
            - '1104:832'
            - '720:1280'
            - '832:1104'
            - '960:960'
    VideoToVideoRequest:
      type: object
      required:
        - model
        - promptVideo
        - promptText
      properties:
        model:
          type: string
          description: >-
            The model to use for video-to-video generation. Currently supports
            the Gen-4 Aleph model.
          enum:
            - gen4_aleph
        promptVideo:
          type: string
          description: >-
            An HTTPS URL or runway:// URI pointing to the input video to
            transform.
        promptText:
          type: string
          description: >-
            A text description of up to 1000 characters guiding the
            transformation of the input video.
          maxLength: 1000
        ratio:
          type: string
          description: >-
            The aspect ratio of the output video.
          enum:
            - '1280:720'
            - '1584:672'
            - '1104:832'
            - '720:1280'
            - '832:1104'
            - '960:960'
    CharacterPerformanceRequest:
      type: object
      required:
        - model
        - character
        - reference
      properties:
        model:
          type: string
          description: >-
            The model to use for character performance. Must be set to act_two.
          enum:
            - act_two
        character:
          type: object
          description: >-
            The character to animate. Accepts either an image or video with a
            URI.
          required:
            - type
            - uri
          properties:
            type:
              type: string
              description: >-
                The media type of the character input.
              enum:
                - image
                - video
            uri:
              type: string
              description: >-
                An HTTPS URL or runway:// URI pointing to the character media.
        reference:
          type: object
          description: >-
            The driving performance reference. Accepts either an image or video
            with a URI that controls the character's movements.
          required:
            - type
            - uri
          properties:
            type:
              type: string
              description: >-
                The media type of the reference input.
              enum:
                - image
                - video
            uri:
              type: string
              description: >-
                An HTTPS URL or runway:// URI pointing to the reference media.
        ratio:
          type: string
          description: >-
            The aspect ratio of the output video.
          enum:
            - '1280:720'
            - '1584:672'
            - '1104:832'
            - '720:1280'
            - '832:1104'
            - '960:960'
    LipSyncRequest:
      type: object
      required:
        - model
        - input
        - audio
      properties:
        model:
          type: string
          description: >-
            The model to use for lip sync generation.
          enum:
            - gen4
            - gen4_turbo
        input:
          type: string
          description: >-
            An HTTPS URL or runway:// URI pointing to the input image or video
            containing the face to animate.
        audio:
          type: string
          description: >-
            An HTTPS URL or runway:// URI pointing to the audio file to sync
            with the face. Supports 28+ languages via eleven_multilingual_v2.
    VideoUpscaleRequest:
      type: object
      required:
        - model
        - videoUri
      properties:
        model:
          type: string
          description: >-
            The model to use for video upscaling.
          enum:
            - upscale_v1
        videoUri:
          type: string
          description: >-
            An HTTPS URL or runway:// URI pointing to the video to upscale.
    FrameInterpolationRequest:
      type: object
      required:
        - model
        - videoUri
      properties:
        model:
          type: string
          description: >-
            The model to use for frame interpolation.
          enum:
            - frame_interpolation_v1
        videoUri:
          type: string
          description: >-
            An HTTPS URL or runway:// URI pointing to the video for frame
            interpolation.
    SoundEffectRequest:
      type: object
      required:
        - model
        - promptText
      properties:
        model:
          type: string
          description: >-
            The model to use for sound effect generation.
          enum:
            - sound_effect_v1
        promptText:
          type: string
          description: >-
            A text description of the desired sound effect.
          maxLength: 1000
        videoUri:
          type: string
          description: >-
            An optional HTTPS URL or runway:// URI pointing to a video to
            generate sound effects for.
        duration:
          type: integer
          description: >-
            The duration of the generated sound effect in seconds.
    TaskCreatedResponse:
      type: object
      properties:
        id:
          type: string
          format: uuid
          description: >-
            The unique identifier for the created task. Use this ID to poll
            the task status endpoint.
    TaskResponse:
      type: object
      properties:
        id:
          type: string
          format: uuid
          description: >-
            The unique identifier of the task.
        status:
          type: string
          description: >-
            The current status of the task.
          enum:
            - PENDING
            - PROCESSING
            - SUCCEEDED
            - FAILED
            - CANCELLED
        createdAt:
          type: string
          format: date-time
          description: >-
            The timestamp when the task was created.
        output:
          type: array
          description: >-
            One or more URLs linking to the generated output media. Only present
            when the task status is SUCCEEDED.
          items:
            type: string
            format: uri
        failure:
          type: string
          description: >-
            A description of the failure reason. Only present when the task
            status is FAILED.
    UploadRequest:
      type: object
      required:
        - filename
      properties:
        filename:
          type: string
          description: >-
            The filename of the file to upload. Must have a valid file extension
            corresponding to a supported media type (image, video, or audio).
    UploadResponse:
      type: object
      properties:
        id:
          type: string
          description: >-
            The identifier for the uploaded file, usable as a runway:// URI in
            generation requests.
        uploadUrl:
          type: string
          format: uri
          description: >-
            The presigned URL to upload the file to via a multipart POST request.
        fields:
          type: object
          description: >-
            A dictionary of form fields to include in the multipart POST request
            when uploading to the uploadUrl.
          additionalProperties:
            type: string
    ErrorResponse:
      type: object
      properties:
        error:
          type: string
          description: >-
            A human-readable error message describing what went wrong.
        code:
          type: string
          description: >-
            A machine-readable error code identifying the type of error.