Introducing Google AI Edge Portal : Benchmark Edge AI at scale. Sign-up to request access during private preview.

MediaPipeTasksVision Framework Reference

RunningMode

  enum 
 RunningMode 
 : 
 UInt 
 , 
 @unchecked 
 Sendable

MediaPipe vision task running mode. A MediaPipe vision task can be run with three different modes: image, video and live stream.

image
The mode for running a mediapipe vision task on single image inputs.
Declaration
Swift

case image = 0
video
The mode for running a mediapipe vision task on the decoded frames of a video.
Declaration
Swift

case video = 1
liveStream
The mode for running a mediapipe vision task on a live stream of input data, such as from the camera.
Declaration
Swift

case liveStream = 2

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License , and code samples are licensed under the Apache 2.0 License . For details, see the Google Developers Site Policies . Java is a registered trademark of Oracle and/or its affiliates.

Last updated 2026-05-28 UTC.

Create a Mobile Website

View Site in Mobile | Classic

Share by: