diff --git a/mediapipe/tasks/web/vision/gesture_recognizer/gesture_recognizer.ts b/mediapipe/tasks/web/vision/gesture_recognizer/gesture_recognizer.ts
index 7441911c1..9ec63b07a 100644
--- a/mediapipe/tasks/web/vision/gesture_recognizer/gesture_recognizer.ts
+++ b/mediapipe/tasks/web/vision/gesture_recognizer/gesture_recognizer.ts
@@ -225,7 +225,9 @@ export class GestureRecognizer extends
 
   /**
    * Performs gesture recognition on the provided single image and waits
-   * synchronously for the response.
+   * synchronously for the response. Only use this method when the
+   * GestureRecognizer is created with running mode `image`.
+   *
    * @param image A single image to process.
    * @return The detected gestures.
    */
@@ -235,7 +237,9 @@ export class GestureRecognizer extends
 
   /**
    * Performs gesture recognition on the provided video frame and waits
-   * synchronously for the response.
+   * synchronously for the response. Only use this method when the
+   * GestureRecognizer is created with running mode `video`.
+   *
    * @param videoFrame A video frame to process.
    * @param timestamp The timestamp of the current frame, in ms.
    * @return The detected gestures.
diff --git a/mediapipe/tasks/web/vision/hand_landmarker/hand_landmarker.ts b/mediapipe/tasks/web/vision/hand_landmarker/hand_landmarker.ts
index 6d69d568c..290f49455 100644
--- a/mediapipe/tasks/web/vision/hand_landmarker/hand_landmarker.ts
+++ b/mediapipe/tasks/web/vision/hand_landmarker/hand_landmarker.ts
@@ -177,7 +177,9 @@ export class HandLandmarker extends VisionTaskRunner<HandLandmarkerResult> {
 
   /**
    * Performs hand landmarks detection on the provided single image and waits
-   * synchronously for the response.
+   * synchronously for the response. Only use this method when the
+   * HandLandmarker is created with running mode `image`.
+   *
    * @param image An image to process.
    * @return The detected hand landmarks.
    */
@@ -187,7 +189,9 @@ export class HandLandmarker extends VisionTaskRunner<HandLandmarkerResult> {
 
   /**
    * Performs hand landmarks detection on the provided video frame and waits
-   * synchronously for the response.
+   * synchronously for the response. Only use this method when the
+   * HandLandmarker is created with running mode `video`.
+   *
    * @param videoFrame A video frame to process.
    * @param timestamp The timestamp of the current frame, in ms.
    * @return The detected hand landmarks.
diff --git a/mediapipe/tasks/web/vision/image_classifier/image_classifier.ts b/mediapipe/tasks/web/vision/image_classifier/image_classifier.ts
index 604795f9f..185ddf9ea 100644
--- a/mediapipe/tasks/web/vision/image_classifier/image_classifier.ts
+++ b/mediapipe/tasks/web/vision/image_classifier/image_classifier.ts
@@ -120,7 +120,8 @@ export class ImageClassifier extends VisionTaskRunner<ImageClassifierResult> {
 
   /**
    * Performs image classification on the provided single image and waits
-   * synchronously for the response.
+   * synchronously for the response. Only use this method when the
+   * ImageClassifier is created with running mode `image`.
    *
    * @param image An image to process.
    * @return The classification result of the image
@@ -131,7 +132,8 @@ export class ImageClassifier extends VisionTaskRunner<ImageClassifierResult> {
 
   /**
    * Performs image classification on the provided video frame and waits
-   * synchronously for the response.
+   * synchronously for the response. Only use this method when the
+   * ImageClassifier is created with running mode `video`.
    *
    * @param videoFrame A video frame to process.
    * @param timestamp The timestamp of the current frame, in ms.
diff --git a/mediapipe/tasks/web/vision/image_embedder/image_embedder.ts b/mediapipe/tasks/web/vision/image_embedder/image_embedder.ts
index 68068db6d..91352e934 100644
--- a/mediapipe/tasks/web/vision/image_embedder/image_embedder.ts
+++ b/mediapipe/tasks/web/vision/image_embedder/image_embedder.ts
@@ -122,10 +122,8 @@ export class ImageEmbedder extends VisionTaskRunner<ImageEmbedderResult> {
 
   /**
    * Performs embedding extraction on the provided single image and waits
-   * synchronously for the response.
-   *
-   * Only use this method when the `useStreamMode` option is not set or
-   * expliclity set to `false`.
+   * synchronously for the response. Only use this method when the
+   * ImageEmbedder is created with running mode `image`.
    *
    * @param image The image to process.
    * @return The classification result of the image
@@ -136,9 +134,8 @@ export class ImageEmbedder extends VisionTaskRunner<ImageEmbedderResult> {
 
   /**
    * Performs embedding extraction on the provided video frame and waits
-   * synchronously for the response.
-   *
-   * Only use this method when the `useStreamMode` option is set to `true`.
+   * synchronously for the response. Only use this method when the
+   * ImageEmbedder is created with running mode `video`.
    *
    * @param imageFrame The image frame to process.
    * @param timestamp The timestamp of the current frame, in ms.
diff --git a/mediapipe/tasks/web/vision/object_detector/object_detector.ts b/mediapipe/tasks/web/vision/object_detector/object_detector.ts
index 0f039acb2..7711c39e9 100644
--- a/mediapipe/tasks/web/vision/object_detector/object_detector.ts
+++ b/mediapipe/tasks/web/vision/object_detector/object_detector.ts
@@ -151,7 +151,9 @@ export class ObjectDetector extends VisionTaskRunner<Detection[]> {
 
   /**
    * Performs object detection on the provided single image and waits
-   * synchronously for the response.
+   * synchronously for the response. Only use this method when the
+   * ObjectDetector is created with running mode `image`.
+   *
    * @param image An image to process.
    * @return The list of detected objects
    */
@@ -161,7 +163,9 @@ export class ObjectDetector extends VisionTaskRunner<Detection[]> {
 
   /**
    * Performs object detection on the provided vidoe frame and waits
-   * synchronously for the response.
+   * synchronously for the response. Only use this method when the
+   * ObjectDetector is created with running mode `video`.
+   *
    * @param videoFrame A video frame to process.
    * @param timestamp The timestamp of the current frame, in ms.
    * @return The list of detected objects