diff --git a/packages/google-cloud-automl/samples/beta/automl_vision_create_model_test.py b/packages/google-cloud-automl/samples/beta/automl_vision_create_model_test.py
new file mode 100644
index 000000000000..bcb6c323e1e4
--- /dev/null
+++ b/packages/google-cloud-automl/samples/beta/automl_vision_create_model_test.py
@@ -0,0 +1,43 @@
+#!/usr/bin/env python
+
+# Copyright 2018 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import datetime
+import os
+
+from google.cloud import automl_v1beta1 as automl
+import pytest
+
+project_id = os.environ["GOOGLE_CLOUD_PROJECT"]
+compute_region = "us-central1"
+
+
+@pytest.mark.skip(reason="creates too many models")
+def test_model_create_status_delete(capsys):
+    # create model
+    client = automl.AutoMlClient()
+    model_name = "test_" + datetime.datetime.now().strftime("%Y%m%d%H%M%S")
+    project_location = client.location_path(project_id, compute_region)
+    my_model = {
+        "display_name": model_name,
+        "dataset_id": "3946265060617537378",
+        "image_classification_model_metadata": {"train_budget": 24},
+    }
+    response = client.create_model(project_location, my_model)
+    operation_name = response.operation.name
+    assert operation_name
+
+    # cancel operation
+    response.cancel()
diff --git a/packages/google-cloud-automl/samples/beta/automl_vision_model.py b/packages/google-cloud-automl/samples/beta/automl_vision_model.py
new file mode 100755
index 000000000000..04aa4c9476e4
--- /dev/null
+++ b/packages/google-cloud-automl/samples/beta/automl_vision_model.py
@@ -0,0 +1,93 @@
+#!/usr/bin/env python
+
+# Copyright 2018 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""This application demonstrates how to perform basic operations on model
+with the Google AutoML Vision API.
+
+For more information, the documentation at
+https://cloud.google.com/vision/automl/docs.
+"""
+
+import argparse
+import os
+
+
+def create_model(
+    project_id, compute_region, dataset_id, model_name, train_budget=24
+):
+    """Create a model."""
+    # [START automl_vision_create_model]
+    # TODO(developer): Uncomment and set the following variables
+    # project_id = 'PROJECT_ID_HERE'
+    # compute_region = 'COMPUTE_REGION_HERE'
+    # dataset_id = 'DATASET_ID_HERE'
+    # model_name = 'MODEL_NAME_HERE'
+    # train_budget = integer amount for maximum cost of model
+
+    from google.cloud import automl_v1beta1 as automl
+
+    client = automl.AutoMlClient()
+
+    # A resource that represents Google Cloud Platform location.
+    project_location = client.location_path(project_id, compute_region)
+
+    # Set model name and model metadata for the image dataset.
+    my_model = {
+        "display_name": model_name,
+        "dataset_id": dataset_id,
+        "image_classification_model_metadata": {"train_budget": train_budget}
+        if train_budget
+        else {},
+    }
+
+    # Create a model with the model metadata in the region.
+    response = client.create_model(project_location, my_model)
+
+    print("Training operation name: {}".format(response.operation.name))
+    print("Training started...")
+
+    # [END automl_vision_create_model]
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(
+        description=__doc__,
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+    subparsers = parser.add_subparsers(dest="command")
+
+    create_model_parser = subparsers.add_parser(
+        "create_model", help=create_model.__doc__
+    )
+    create_model_parser.add_argument("dataset_id")
+    create_model_parser.add_argument("model_name")
+    create_model_parser.add_argument(
+        "train_budget", type=int, nargs="?", default=0
+    )
+
+    project_id = os.environ["PROJECT_ID"]
+    compute_region = os.environ["REGION_NAME"]
+
+    args = parser.parse_args()
+
+    if args.command == "create_model":
+        create_model(
+            project_id,
+            compute_region,
+            args.dataset_id,
+            args.model_name,
+            args.train_budget,
+        )
diff --git a/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/Dockerfile b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/Dockerfile
new file mode 100644
index 000000000000..d447bcc57544
--- /dev/null
+++ b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/Dockerfile
@@ -0,0 +1,29 @@
+# Copyright 2019 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+ARG TF_SERVING_IMAGE_TAG
+FROM tensorflow/serving:${TF_SERVING_IMAGE_TAG}
+
+ENV GCS_READ_CACHE_MAX_STALENESS 300
+ENV GCS_STAT_CACHE_MAX_AGE 300
+ENV GCS_MATCHING_PATHS_CACHE_MAX_AGE 300
+
+EXPOSE 8500
+EXPOSE 8501
+ENTRYPOINT /usr/bin/tensorflow_model_server \
+              --port=8500 \
+              --rest_api_port=8501 \
+              --model_base_path=/tmp/mounted_model/ \
+              --tensorflow_session_parallelism=0 \
+              --file_system_poll_wait_seconds=31540000
diff --git a/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/README.md b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/README.md
new file mode 100644
index 000000000000..da8cc50ed66f
--- /dev/null
+++ b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/README.md
@@ -0,0 +1,78 @@
+# AutoML Vision Edge Container Prediction
+
+This is an example to show how to predict with AutoML Vision Edge Containers.
+The test (automl_vision_edge_container_predict_test.py) shows an automatical way
+to run the prediction.
+
+If you want to try the test manually with a sample model, please install
+[gsutil tools](https://cloud.google.com/storage/docs/gsutil_install) and
+[Docker CE](https://docs.docker.com/install/) first, and then follow the steps
+below. All the following instructions with commands assume you are in this
+folder with system variables as
+
+```bash
+$ CONTAINER_NAME=AutomlVisionEdgeContainerPredict
+$ PORT=8505
+```
+
++   Step 1. Pull the Docker image.
+
+```bash
+# This is a CPU TFServing 1.14.0 with some default settings compiled from
+# https://hub.docker.com/r/tensorflow/serving.
+$ DOCKER_GCS_DIR=gcr.io/cloud-devrel-public-resources
+$ CPU_DOCKER_GCS_PATH=${DOCKER_GCS_DIR}/gcloud-container-1.14.0:latest
+$ sudo docker pull ${CPU_DOCKER_GCS_PATH}
+```
+
++   Step 2. Get a sample saved model.
+
+```bash
+$ MODEL_GCS_DIR=gs://cloud-samples-data/vision/edge_container_predict
+$ SAMPLE_SAVED_MODEL=${MODEL_GCS_DIR}/saved_model.pb
+$ mkdir model_path
+$ YOUR_MODEL_PATH=$(realpath model_path)
+$ gsutil -m cp ${SAMPLE_SAVED_MODEL} ${YOUR_MODEL_PATH}
+```
+
++   Step 3. Run the Docker container.
+
+```bash
+$ sudo docker run --rm --name ${CONTAINER_NAME} -p ${PORT}:8501 -v \
+    ${YOUR_MODEL_PATH}:/tmp/mounted_model/0001 -t ${CPU_DOCKER_GCS_PATH}
+```
+
++   Step 4. Send a prediction request.
+
+```bash
+$ python automl_vision_edge_container_predict.py --image_file_path=./test.jpg \
+    --image_key=1 --port_number=${PORT}
+```
+
+The outputs are
+
+```
+{
+    'predictions':
+    [
+        {
+            'scores': [0.0914393, 0.458942, 0.027604, 0.386767, 0.0352474],
+            labels': ['daisy', 'dandelion', 'roses', 'sunflowers', 'tulips'],
+            'key': '1'
+        }
+    ]
+}
+```
+
++   Step 5. Stop the container.
+
+```bash
+sudo docker stop ${CONTAINER_NAME}
+```
+
+Note: The docker image is uploaded with the following command.
+
+```bash
+gcloud builds --project=cloud-devrel-public-resources \
+  submit --config cloudbuild.yaml
+```
diff --git a/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/automl_vision_edge_container_predict.py b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/automl_vision_edge_container_predict.py
new file mode 100644
index 000000000000..a5b1ed9f2d44
--- /dev/null
+++ b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/automl_vision_edge_container_predict.py
@@ -0,0 +1,80 @@
+#!/usr/bin/env python
+
+# Copyright 2019 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+r"""This is an example to call REST API from TFServing docker containers.
+
+Examples:
+    python automl_vision_edge_container_predict.py \
+    --image_file_path=./test.jpg --image_key=1 --port_number=8051
+
+"""
+
+import argparse
+# [START automl_vision_edge_container_predict]
+import base64
+import io
+import json
+
+import requests
+
+
+def container_predict(image_file_path, image_key, port_number=8501):
+    """Sends a prediction request to TFServing docker container REST API.
+
+    Args:
+        image_file_path: Path to a local image for the prediction request.
+        image_key: Your chosen string key to identify the given image.
+        port_number: The port number on your device to accept REST API calls.
+    Returns:
+        The response of the prediction request.
+    """
+
+    with io.open(image_file_path, 'rb') as image_file:
+        encoded_image = base64.b64encode(image_file.read()).decode('utf-8')
+
+    # The example here only shows prediction with one image. You can extend it
+    # to predict with a batch of images indicated by different keys, which can
+    # make sure that the responses corresponding to the given image.
+    instances = {
+            'instances': [
+                    {'image_bytes': {'b64': str(encoded_image)},
+                     'key': image_key}
+            ]
+    }
+
+    # This example shows sending requests in the same server that you start
+    # docker containers. If you would like to send requests to other servers,
+    # please change localhost to IP of other servers.
+    url = 'http://localhost:{}/v1/models/default:predict'.format(port_number)
+
+    response = requests.post(url, data=json.dumps(instances))
+    print(response.json())
+    # [END automl_vision_edge_container_predict]
+    return response.json()
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--image_file_path', type=str)
+    parser.add_argument('--image_key', type=str, default='1')
+    parser.add_argument('--port_number', type=int, default=8501)
+    args = parser.parse_args()
+
+    container_predict(args.image_file_path, args.image_key, args.port_number)
+
+
+if __name__ == '__main__':
+    main()
diff --git a/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/automl_vision_edge_container_predict_test.py b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/automl_vision_edge_container_predict_test.py
new file mode 100644
index 000000000000..327513c034b7
--- /dev/null
+++ b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/automl_vision_edge_container_predict_test.py
@@ -0,0 +1,108 @@
+#!/usr/bin/env python
+
+# Copyright 2019 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""Tests for automl_vision_edge_container_predict.
+
+The test will automatically start a container with a sample saved_model.pb,
+send a request with one image, verify the response and delete the started
+container.
+
+If you want to try the test, please install
+[gsutil tools](https://cloud.google.com/storage/docs/gsutil_install) and
+[Docker CE](https://docs.docker.com/install/) first.
+
+Examples:
+sudo python -m pytest automl_vision_edge_container_predict_test.py
+"""
+
+import os
+import subprocess
+import tempfile
+import time
+
+import pytest
+
+import automl_vision_edge_container_predict as predict  # noqa
+
+
+IMAGE_FILE_PATH = os.path.join(os.path.dirname(__file__), 'test.jpg')
+# The cpu docker gcs path is from 'Edge container tutorial'.
+CPU_DOCKER_GCS_PATH = '{}'.format(
+  'gcr.io/cloud-devrel-public-resources/gcloud-container-1.14.0:latest')
+# The path of a sample saved model.
+SAMPLE_SAVED_MODEL = '{}'.format(
+    'gs://cloud-samples-data/vision/edge_container_predict/saved_model.pb')
+# Container Name.
+NAME = 'AutomlVisionEdgeContainerPredictTest'
+# Port Number.
+PORT_NUMBER = 8505
+
+
+@pytest.fixture
+def edge_container_predict_server_port():
+    # set up
+    # Pull the CPU docker.
+    subprocess.check_output(
+        ['docker', 'pull', CPU_DOCKER_GCS_PATH],
+        env={'DOCKER_API_VERSION': '1.38'})
+
+    if os.environ.get('TRAMPOLINE_VERSION'):
+        # Use /tmp
+        model_path = tempfile.TemporaryDirectory()
+    else:
+        # Use project directory with Trampoline V1.
+        model_path = tempfile.TemporaryDirectory(dir=os.path.dirname(__file__))
+    print("Using model_path: {}".format(model_path))
+    # Get the sample saved model.
+    subprocess.check_output(
+        ['gsutil', '-m', 'cp', SAMPLE_SAVED_MODEL, model_path.name])
+
+    # Start the CPU docker.
+    subprocess.Popen(['docker', 'run', '--rm', '--name', NAME, '-v',
+                      model_path.name + ':/tmp/mounted_model/0001', '-p',
+                      str(PORT_NUMBER) + ':8501', '-t',
+                      CPU_DOCKER_GCS_PATH],
+                     env={'DOCKER_API_VERSION': '1.38'})
+    # Sleep a few seconds to wait for the container running.
+    time.sleep(10)
+
+    yield PORT_NUMBER
+
+    # tear down
+    # Stop the container.
+    subprocess.check_output(
+        ['docker', 'stop', NAME], env={'DOCKER_API_VERSION': '1.38'})
+    # Remove the docker image.
+    subprocess.check_output(
+        ['docker', 'rmi', CPU_DOCKER_GCS_PATH],
+        env={'DOCKER_API_VERSION': '1.38'})
+    # Remove the temporery directory.
+    model_path.cleanup()
+
+
+def test_edge_container_predict(capsys, edge_container_predict_server_port):
+    # If you send requests with one image each time, the key value does not
+    # matter. If you send requests with multiple images, please used different
+    # keys to indicated different images, which can make sure that the
+    # responses corresponding to the given image.
+    image_key = '1'
+    # Send a request.
+    response = predict.container_predict(
+            IMAGE_FILE_PATH, image_key, PORT_NUMBER)
+    # Verify the response.
+    assert 'predictions' in response
+    assert 'key' in response['predictions'][0]
+    assert image_key == response['predictions'][0]['key']
diff --git a/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/cloudbuild.yaml b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/cloudbuild.yaml
new file mode 100644
index 000000000000..5b0760254b4b
--- /dev/null
+++ b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/cloudbuild.yaml
@@ -0,0 +1,26 @@
+# Copyright 2020 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+timeout: 3600s
+
+steps:
+    - name: 'gcr.io/cloud-builders/docker'
+      args: ['build', '-t', 'gcr.io/$PROJECT_ID/gcloud-container-1.14.0',
+             '--build-arg', 'TF_SERVING_IMAGE_TAG=1.14.0', '.']
+    - name: 'gcr.io/cloud-builders/docker'
+      args: ['build', '-t', 'gcr.io/$PROJECT_ID/gcloud-container-1.14.0-gpu',
+             '--build-arg', 'TF_SERVING_IMAGE_TAG=1.14.0-gpu', '.']
+images:
+     - 'gcr.io/$PROJECT_ID/gcloud-container-1.14.0'
+     - 'gcr.io/$PROJECT_ID/gcloud-container-1.14.0-gpu'
diff --git a/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/test.jpg b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/test.jpg
new file mode 100644
index 000000000000..4873e8d1b523
Binary files /dev/null and b/packages/google-cloud-automl/samples/vision_edge/edge_container_predict/test.jpg differ
diff --git a/packages/google-cloud-automl/samples/vision_edge/resources/test.png b/packages/google-cloud-automl/samples/vision_edge/resources/test.png
new file mode 100644
index 000000000000..653342a46e58
Binary files /dev/null and b/packages/google-cloud-automl/samples/vision_edge/resources/test.png differ