feat: Add ImageResizeMode for GenerateVideos

matthew29tang · copybara-github · commit 317d2af040ad · 2026-04-24T11:37:45.000-07:00
PiperOrigin-RevId: 905140128
diff --git a/google/genai/models.py b/google/genai/models.py
@@ -2163,6 +2163,9 @@ def _GenerateVideosConfig_to_mldev(
         parent_object, ['webhookConfig'], getv(from_object, ['webhook_config'])
     )
 
+  if getv(from_object, ['resize_mode']) is not None:
+    raise ValueError('resize_mode parameter is not supported in Gemini API.')
+
   return to_object
 
 
@@ -2295,6 +2298,13 @@ def _GenerateVideosConfig_to_vertex(
         ' Platform.'
     )
 
+  if getv(from_object, ['resize_mode']) is not None:
+    setv(
+        parent_object,
+        ['parameters', 'resizeMode'],
+        getv(from_object, ['resize_mode']),
+    )
+
   return to_object
 
 
diff --git a/google/genai/tests/models/test_generate_videos.py b/google/genai/tests/models/test_generate_videos.py
@@ -335,6 +335,8 @@ def test_text_and_image_to_video_poll(client):
       image=GCS_IMAGE if client.vertexai else LOCAL_IMAGE,
       config=types.GenerateVideosConfig(
           output_gcs_uri=OUTPUT_GCS_URI if client.vertexai else None,
+          resize_mode=(types.ImageResizeMode.CROP
+                       if client.vertexai else None),
       ),
   )
   while not operation.done:
diff --git a/google/genai/types.py b/google/genai/types.py
@@ -1005,6 +1005,17 @@ class VideoCompressionQuality(_common.CaseInSensitiveEnum):
       with a larger file size."""
 
 
+class ImageResizeMode(_common.CaseInSensitiveEnum):
+  """Resize mode for the image input for video generation."""
+
+  CROP = 'CROP'
+  """Crop the image to fit the correct aspect ratio (so we lose parts
+      of the image in the process)."""
+  PAD = 'PAD'
+  """Pad the image to fit the correct aspect ratio (so we don't lose
+      any parts of the image in the process)."""
+
+
 class TuningMethod(_common.CaseInSensitiveEnum):
   """Enum representing the tuning method."""
 
@@ -11233,6 +11244,10 @@ class GenerateVideosConfig(_common.BaseModel):
       description="""Webhook configuration for receiving notifications when the
       video generation operation completes.""",
   )
+  resize_mode: Optional[ImageResizeMode] = Field(
+      default=None,
+      description="""Resize mode of the image input for video generation.""",
+  )
 
 
 class GenerateVideosConfigDict(TypedDict, total=False):
@@ -11309,6 +11324,9 @@ class GenerateVideosConfigDict(TypedDict, total=False):
   """Webhook configuration for receiving notifications when the
       video generation operation completes."""
 
+  resize_mode: Optional[ImageResizeMode]
+  """Resize mode of the image input for video generation."""
+
 
 GenerateVideosConfigOrDict = Union[
     GenerateVideosConfig, GenerateVideosConfigDict

Original file line number	Diff line number	Diff line change
`@@ -335,6 +335,8 @@ def test_text_and_image_to_video_poll(client):`
`335`	`335`	`image=GCS_IMAGE if client.vertexai else LOCAL_IMAGE,`
`336`	`336`	`config=types.GenerateVideosConfig(`
`337`	`337`	`output_gcs_uri=OUTPUT_GCS_URI if client.vertexai else None,`
	`338`	`+ resize_mode=(types.ImageResizeMode.CROP`
	`339`	`+ if client.vertexai else None),`
`338`	`340`	`),`
`339`	`341`	`)`
`340`	`342`	`while not operation.done:`