Image Frames
Frame types for handling image data and visual interactions in Pipecat
ImageRawFrame
Base class for all image-related frames.
Properties
Raw image data
Image dimensions as (width, height)
Image format identifier (e.g., “jpeg”, “png”)
Methods
InputImageRawFrame
Base class for images coming from input sources.
OutputImageRawFrame
Base class for images intended for display or output.
UserImageRawFrame
Frame for images associated with specific users.
Properties
Inherits ImageRawFrame properties, plus:
Identifier of the user associated with the image
VisionImageRawFrame
Frame for images that require AI vision processing.
Properties
Inherits ImageRawFrame properties, plus:
Text prompt or description associated with the image
URLImageRawFrame
Frame for images referenced by URL.
Properties
Inherits OutputImageRawFrame properties, plus:
URL where the image can be accessed
SpriteFrame
Frame for animated sprite sequences.
Properties
List of image frames comprising the animation
Methods
Usage Examples
Creating Image Frames
Common Pipeline Usage
Frame Flow
Notes
- Image data should be in raw bytes format
- Supported formats depend on your pipeline’s image processing capabilities
- Sprite frames are played at the rate specified by transport’s
camera_out_framerate
- Vision frames may be used with multimodal LLMs for image understanding