diff --git a/.github/workflows/test.yaml b/.github/workflows/test.yaml
index 0b319b8b..0482956d 100644
--- a/.github/workflows/test.yaml
+++ b/.github/workflows/test.yaml
@@ -30,6 +30,9 @@ jobs:
         with:
           environment-file: environment_cpu.yaml
 
+      # Setup Qt libraries for GUI testing
+      - uses: tlambert03/setup-qt-libs@v1
+
       - name: Install pytest
         shell: bash -l {0}
         run: |
@@ -37,8 +40,10 @@ jobs:
           python -m pip install pytest-cov
 
       - name: Install package
-        shell: bash -l {0}
-        run: pip install --no-deps -e .
+        uses: aganders3/headless-gui@v1
+        with:
+          shell: bash -l {0}
+          run: pip install --no-deps -e .
 
       - name: Run tests
         shell: bash -l {0}
diff --git a/.github/workflows/test_gui.yaml b/.github/workflows/test_gui.yaml
new file mode 100644
index 00000000..e795834f
--- /dev/null
+++ b/.github/workflows/test_gui.yaml
@@ -0,0 +1,127 @@
+# Our minimal suite of tests that run on each pull request
+name: PR Test
+
+on:
+  pull_request:
+    branches:
+      - main
+      - dev
+      - "v*x"
+
+concurrency:
+  group: test-${{ github.ref }}
+  cancel-in-progress: true
+
+env:
+  COLUMNS: 120
+
+jobs:
+  test:
+    name: ${{ matrix.platform }} ${{ matrix.python }} ${{ matrix.toxenv || matrix.backend }} ${{ matrix.MIN_REQ && 'min_req' }}
+    runs-on: ${{ matrix.platform }}
+    timeout-minutes: 40
+    strategy:
+      fail-fast: false
+      matrix:
+        platform: [ubuntu-latest]
+        python: ["3.8", "3.9", "3.10", "3.11"]
+        backend: [pyqt5, pyside2]
+        exclude:
+          - python: '3.11'
+            backend: pyside2
+        include:
+          # Windows py38
+          - python: 3.8
+            platform: windows-latest
+            backend: pyqt5
+          - python: 3.8
+            platform: windows-latest
+            backend: pyside2
+          - python: 3.9
+            platform: macos-latest
+            backend: pyqt5
+          # minimum specified requirements
+          - python: 3.8
+            platform: ubuntu-20.04
+            backend: pyqt5
+            MIN_REQ: 1
+          # test without any Qt backends
+          - python: 3.8
+            platform: ubuntu-20.04
+            backend: headless
+          - python: 3.9
+            platform: ubuntu-latest
+            backend: pyqt6
+          - python: 3.9
+            platform: ubuntu-latest
+            backend: pyside6
+          # pyside 6
+          - python: '3.10'
+            platform: ubuntu-latest
+            backend: pyside6
+          - python: '3.11'
+            platform: ubuntu-latest
+            backend: pyside6
+
+    steps:
+      - name: Cancel Previous Runs
+        uses: styfle/cancel-workflow-action@0.11.0
+        with:
+          access_token: ${{ github.token }}
+
+      - uses: actions/checkout@v4
+
+      - name: Set up Python ${{ matrix.python }}
+        uses: actions/setup-python@v4
+        with:
+          python-version: ${{ matrix.python }}
+          cache: "pip"
+          cache-dependency-path: setup.cfg
+
+      - uses: tlambert03/setup-qt-libs@v1
+
+      # strategy borrowed from vispy for installing opengl libs on windows
+      - name: Install Windows OpenGL
+        if: runner.os == 'Windows'
+        run: |
+          git clone --depth 1 https://github.com/pyvista/gl-ci-helpers.git
+          powershell gl-ci-helpers/appveyor/install_opengl.ps1
+          if (Test-Path -Path "C:\Windows\system32\opengl32.dll" -PathType Leaf) {Exit 0} else {Exit 1}
+
+      # tox and tox-gh-actions will take care of the "actual" installation
+      # of python dependendencies into a virtualenv.  see tox.ini for more
+      - name: Install dependencies
+        run: |
+          pip install --upgrade pip
+          pip install setuptools tox tox-gh-actions tox-min-req
+
+      # here we pass off control of environment creation and running of tests to tox
+      # tox-gh-actions, installed above, helps to convert environment variables into
+      # tox "factors" ... limiting the scope of what gets tested on each platform
+      # for instance, on ubuntu-latest with python 3.8, it would be equivalent to this command:
+      # `tox -e py38-linux-pyqt,py38-linux-pyside`
+      # see tox.ini for more
+      - name: Test with tox
+        # the longest is macos-latest 3.9 pyqt5 at ~30 minutes.
+        timeout-minutes: 40
+        uses: aganders3/headless-gui@v1
+        with:
+          run: python -m tox
+        env:
+          PLATFORM: ${{ matrix.platform }}
+          BACKEND: ${{ matrix.backend }}
+          TOXENV: ${{ matrix.toxenv }}
+          NUMPY_EXPERIMENTAL_ARRAY_FUNCTION: ${{ matrix.MIN_REQ || 1 }}
+          PYVISTA_OFF_SCREEN: True
+          MIN_REQ: ${{ matrix.MIN_REQ }}
+          FORCE_COLOR: 1
+          PIP_CONSTRAINT: resources/constraints/constraints_py${{ matrix.python }}${{ matrix.MIN_REQ && '_min_req' }}.txt
+      - uses: actions/upload-artifact@v3
+        with:
+          name: upload pytest timing reports as json
+          path: |
+            ./report-*.json
+
+      - name: Coverage
+        uses: codecov/codecov-action@v3
+
diff --git a/test/test_gui.py b/test/test_gui.py
new file mode 100644
index 00000000..43944d40
--- /dev/null
+++ b/test/test_gui.py
@@ -0,0 +1,195 @@
+import numpy as np
+import skimage.data
+from micro_sam.sam_annotator import annotator_2d, annotator_3d
+from micro_sam.sam_annotator.annotator_2d import _initialize_viewer, _segment_widget, _autosegment_widget
+from micro_sam.sam_annotator.util import _clear_widget, _commit_segmentation_widget
+
+
+def _check_layer_initialization(viewer):
+    """Utility function to check the initial layer setup is correct."""
+    assert len(viewer.layers) == 6
+    expected_layer_names = ['raw', 'auto_segmentation', 'committed_objects', 'current_object', 'point_prompts', 'prompts']
+    for layername in expected_layer_names:
+        assert layername in viewer.layers
+    # Check layers are empty before beginning tests
+    np.testing.assert_equal(viewer.layers["auto_segmentation"].data, 0)
+    np.testing.assert_equal(viewer.layers["current_object"].data, 0)
+    np.testing.assert_equal(viewer.layers["committed_objects"].data, 0)
+    np.testing.assert_equal(viewer.layers["point_prompts"].data, 0)
+    assert viewer.layers["prompts"].data == []  # shape data is list, not numpy array
+
+
+def test_annotator_2d_amg(make_napari_viewer_proxy, tmp_path):
+    """Integration test for annotator_2d widget with automatic mask generation.
+
+    * Creates 2D image embedding
+    * Opens annotator_2d widget in napari
+    * Test automatic mask generation
+    """
+    model_type = "vit_b"
+    embedding_path = tmp_path / "test-embedding.zarr"
+    # example data - a basic checkerboard pattern
+    image = np.zeros((16,16,16))
+    image[:8,:8,:8] = 1
+    image[8:,8:,8:] = 1
+
+    viewer = make_napari_viewer_proxy()
+    viewer = _initialize_viewer(image, None, None, None)  # TODO: fix hacky workaround
+    # test generating image embedding, then adding micro-sam dock widgets to the GUI
+    viewer = annotator_2d(
+        image,
+        embedding_path,
+        show_embeddings=False,
+        model_type=model_type,
+        v=viewer,
+        return_viewer=True
+    )
+    _check_layer_initialization(viewer)
+    # ========================================================================
+    # # Automatic mask generation
+    # _autosegment_widget(v=viewer, min_object_size=30)
+    # # We expect four segmentation regions to be identified
+    # expected_segmentation_label_ids = np.array([0,1,2,3])
+    # np.testing.assert_equal(np.unique(viewer.layers["auto_segmentation"].data),
+    #                         expected_segmentation_label_ids)
+    viewer.close()  # must close the viewer at the end of tests
+
+
+def test_annotator_3d(make_napari_viewer_proxy, tmp_path):
+    """Integration test for annotator_2d widget with automatic mask generation.
+
+    * Creates 2D image embedding
+    * Opens annotator_2d widget in napari
+    * Test automatic mask generation
+    """
+    model_type = "vit_b"
+    embedding_path = tmp_path / "test-embedding.zarr"
+    # example data - a basic checkerboard pattern
+    image = np.zeros((16,16))
+    image[:8,:8] = 1
+    image[8:,8:] = 1
+
+    viewer = make_napari_viewer_proxy()
+    viewer = _initialize_viewer(image, None, None, None)  # TODO: fix hacky workaround
+    # test generating image embedding, then adding micro-sam dock widgets to the GUI
+    viewer = annotator_3d(
+        image,
+        embedding_path,
+        show_embeddings=False,
+        model_type=model_type,
+        v=viewer,
+        return_viewer=True
+    )
+    _check_layer_initialization(viewer)
+    # ========================================================================
+    # # Automatic mask generation
+    # _autosegment_widget(v=viewer, min_object_size=30)
+    # # We expect four segmentation regions to be identified
+    # expected_segmentation_label_ids = np.array([0,1,2,3])
+    # np.testing.assert_equal(np.unique(viewer.layers["auto_segmentation"].data),
+    #                         expected_segmentation_label_ids)
+    viewer.close()  # must close the viewer at the end of tests
+
+
+# def test_annotator_2d(make_napari_viewer_proxy, tmp_path):
+#     """Integration test for annotator_2d widget.
+
+#     * Creates 2D image embedding
+#     * Opens annotator_2d widget in napari
+#     * Test point prompts (add points, segment object, clear, and commit)
+#     * Test box prompt (add rectangle prompt, segment object, clear, and commit)
+#     ...
+#     """
+#     model_type = "vit_b"
+#     image = skimage.data.camera()
+#     embedding_path = tmp_path / "test-embedding.zarr"
+
+#     viewer = make_napari_viewer_proxy()
+#     viewer = _initialize_viewer(image, None, None, None)  # TODO: fix hacky workaround
+#     # test generating image embedding, then adding micro-sam dock widgets to the GUI
+#     viewer = annotator_2d(
+#         image,
+#         embedding_path,
+#         show_embeddings=False,
+#         model_type=model_type,
+#         v=viewer,
+#         return_viewer=True
+#     )
+#     _assert_initialization_is_correct(viewer)
+
+#     # ========================================================================
+#     # TEST POINT PROMPTS
+#     # Add three points in the sky region of the camera image
+#     sky_point_prompts = np.array([[70, 80],[50, 320],[80, 470 ]])
+#     viewer.layers["point_prompts"].data = sky_point_prompts
+
+#     # Segment sky region of image
+#     _segment_widget(v=viewer)  # segment slice
+#     # We expect all of the first 50 rows should be identified as sky,
+#     assert (viewer.layers["current_object"].data[0:50,:] == 1).all()
+#     # We also expect roughly 25% of the image to be sky
+#     sky_segmentation = np.copy(viewer.layers["current_object"].data)
+#     segmented_pixel_percentage = (np.sum(sky_segmentation == 1) / image.size) * 100
+#     assert segmented_pixel_percentage > 25
+#     assert segmented_pixel_percentage < 30
+
+#     # Clear segmentation current object and prompts
+#     _clear_widget(v=viewer)
+#     np.testing.assert_equal(viewer.layers["current_object"].data, 0)
+#     np.testing.assert_equal(viewer.layers["point_prompts"].data, 0)
+#     assert viewer.layers["prompts"].data == []  # shape data is list, not numpy array
+
+#     # Repeat segmentation and commit segmentation result
+#     viewer.layers["point_prompts"].data = sky_point_prompts
+#     _segment_widget(v=viewer)  # segment slice
+#     np.testing.assert_equal(sky_segmentation, viewer.layers["current_object"].data)
+#     # Commit segmentation
+#     _commit_segmentation_widget(v=viewer)
+#     np.testing.assert_equal(sky_segmentation, viewer.layers["committed_objects"].data)
+
+#     # ========================================================================
+#     # TEST BOX PROMPTS
+#     # Add rechangle bounding box prompt
+#     camera_bounding_box_prompt = np.array([[139, 254],[139, 324],[183, 324],[183, 254]])
+#     viewer.layers["prompts"].data = [camera_bounding_box_prompt]
+#     # Segment slice
+#     _segment_widget(v=viewer)  # segment slice
+#     # Check segmentation results
+#     camera_segmentation = np.copy(viewer.layers["current_object"].data)
+#     segmented_pixels = np.sum(camera_segmentation == 1)
+#     assert segmented_pixels > 2500  # we expect roughly 2770 pixels
+#     assert segmented_pixels < 3000  # we expect roughly 2770 pixels
+#     assert (camera_segmentation[150:175,275:310] == 1).all()  # small patch which should definitely be inside segmentation
+
+#     # Clear segmentation current object and prompts
+#     _clear_widget(v=viewer)
+#     np.testing.assert_equal(viewer.layers["current_object"].data, 0)
+#     np.testing.assert_equal(viewer.layers["point_prompts"].data, 0)
+#     assert viewer.layers["prompts"].data == []  # shape data is list, not numpy array
+
+#     # Repeat segmentation and commit segmentation result
+#     viewer.layers["prompts"].data = [camera_bounding_box_prompt]
+#     _segment_widget(v=viewer)  # segment slice
+#     np.testing.assert_equal(camera_segmentation, viewer.layers["current_object"].data)
+#     # Commit segmentation
+#     _commit_segmentation_widget(v=viewer)
+#     committed_objects = viewer.layers["committed_objects"].data
+#     # We expect two committed objects
+#     # label id 1: sky segmentation
+#     # label id 2: camera segmentation
+#     np.testing.assert_equal(np.unique(committed_objects), np.array([0, 1, 2]))
+#     np.testing.assert_equal(committed_objects == 2, camera_segmentation == 1)
+
+#     # ========================================================================
+#     viewer.close()  # must close the viewer at the end of tests
+
+
+
+
+
+# def test_something_else(make_napari_viewer_proxy):
+#     viewer = make_napari_viewer_proxy()
+#     # carry on with your test
+#     image = skimage.data.brick()
+#     viewer.add_image(image, name="raw")
+#     viewer.close()