title: Pika API Vocabulary
description: Domain vocabulary for the Pika AI video generation API, including terms, concepts, and definitions used across endpoints.
url: https://pika.art/api
version: '1.0'
created: '2026-06-12'

terms:

  - term: prompt
    label: Prompt
    definition: >
      A natural language text description that guides the AI model in generating the desired
      video content, motion style, camera movement, and visual aesthetic.
    domain: Input
    type: string

  - term: negative_prompt
    label: Negative Prompt
    definition: >
      A text description of visual qualities, artifacts, or content to avoid in the generated
      video. Helps steer the model away from undesirable outputs.
    domain: Input
    type: string

  - term: image_url
    label: Image URL
    definition: >
      The URL of a static image to be used as the first frame in image-to-video generation.
      The model animates this image based on the accompanying prompt.
    domain: Input
    type: string (URI)

  - term: resolution
    label: Resolution
    definition: >
      The output video resolution. Pika 2.2 supports 720p (standard) and 1080p (high definition).
      Higher resolution results in larger file sizes and higher per-clip cost.
    domain: Input
    type: enum
    values:
      - 720p
      - 1080p

  - term: duration
    label: Duration
    definition: >
      The length of the generated video clip in seconds. Pika 2.2 supports 5-second and
      10-second clips.
    domain: Input
    type: integer
    values:
      - 5
      - 10

  - term: aspect_ratio
    label: Aspect Ratio
    definition: >
      The width-to-height ratio of the generated video frame. Supports landscape (16:9),
      portrait (9:16), square (1:1), and various social media formats.
    domain: Input
    type: enum
    values:
      - "16:9"
      - "9:16"
      - "1:1"
      - "4:5"
      - "5:4"
      - "3:2"
      - "2:3"

  - term: seed
    label: Seed
    definition: >
      An integer value used to initialize the random number generator for reproducible
      video generation results. Identical seeds with identical inputs produce identical outputs.
    domain: Input
    type: integer (nullable)

  - term: request_id
    label: Request ID
    definition: >
      A unique identifier assigned to each generation request when submitted to the fal.ai
      queue. Used to poll status, retrieve results, or cancel the request.
    domain: Queue
    type: string (UUID)

  - term: queue_position
    label: Queue Position
    definition: >
      The current position of a request in the processing queue. A value of 0 indicates
      the request is next to be processed.
    domain: Queue
    type: integer

  - term: status
    label: Status
    definition: >
      The current processing state of a queued generation request.
    domain: Queue
    type: enum
    values:
      - IN_QUEUE
      - IN_PROGRESS
      - COMPLETED

  - term: text-to-video
    label: Text-to-Video
    definition: >
      A video generation mode where only a text prompt is used to create a new video clip.
      Pika 2.2 uses a two-stage pipeline internally: text-to-image followed by image-to-video.
    domain: Feature
    type: concept

  - term: image-to-video
    label: Image-to-Video
    definition: >
      A video generation mode that takes a static image as the first frame and animates it
      based on a text prompt. Pika 2.2 animates the image with realistic motion and specified
      camera movement.
    domain: Feature
    type: concept

  - term: FAL_KEY
    label: FAL Key
    definition: >
      The API authentication key issued by fal.ai, used to authorize requests to the Pika
      2.2 API endpoints. Should be passed in the Authorization header as "Key YOUR_FAL_KEY".
    domain: Authentication
    type: string

  - term: pikascenes
    label: Pikascenes
    definition: >
      A Pika platform feature for composing multi-scene video sequences by combining
      multiple generated or uploaded clips into a coherent narrative.
    domain: Feature
    type: concept

  - term: pikaframes
    label: Pikaframes
    definition: >
      A Pika platform feature for keyframe interpolation — generating smooth video transitions
      between two specified keyframe images.
    domain: Feature
    type: concept

  - term: pikaffects
    label: Pikaffects
    definition: >
      A Pika platform feature that applies specialized visual transformations and motion
      effects to existing images or video clips.
    domain: Feature
    type: concept

  - term: pikaswaps
    label: Pikaswaps
    definition: >
      A Pika platform feature for replacing or swapping elements within a video using
      AI-driven content generation.
    domain: Feature
    type: concept

  - term: pikadditions
    label: Pikadditions
    definition: >
      A Pika platform feature for adding new visual elements or objects into an existing
      video scene using generative AI.
    domain: Feature
    type: concept