feat(genai): model and SDK updates for 2025 May (#13381)

* feat(genai): update to use latest SDK version

* feat(genai): update to use global endpoint as location

* fix(genai): copyright year

All genai samples are created only in 2025.

* feat(genai): update to use gemini 2.5 flash

* fix(genai): lint error

* fix(genai):  error

* fix(genai): lint error

* fix(genai): No global endpoint support for Model optimiser

* fix(genai): rollback model updates for function calling

Author: msampathkumar

genai/batch_prediction/requirements.txt

@@ -1 +1 @@
-google-genai==1.7.0
+google-genai==1.16.1

genai/bounding_box/boundingbox_with_txt_img.py

@@ -94,7 +94,7 @@ def plot_bounding_boxes(image_uri: str, bounding_boxes: list[BoundingBox]) -> No
     image_uri = "https://storage.googleapis.com/generativeai-downloads/images/socks.jpg"
 
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents=[
             Part.from_uri(
                 file_uri=image_uri,

genai/bounding_box/requirements.txt

@@ -1,2 +1,2 @@
-google-genai==1.7.0
+google-genai==1.16.1
 pillow==11.1.0

genai/bounding_box/test_bounding_box_examples.py

@@ -21,7 +21,7 @@
 import boundingbox_with_txt_img
 
 os.environ["GOOGLE_GENAI_USE_VERTEXAI"] = "True"
-os.environ["GOOGLE_CLOUD_LOCATION"] = "us-central1"
+os.environ["GOOGLE_CLOUD_LOCATION"] = "global"  # "us-central1"
 # The project name is included in the CICD pipeline
 # os.environ['GOOGLE_CLOUD_PROJECT'] = "add-your-project-name"
 

genai/content_cache/contentcache_create_with_txt_gcs_pdf.py

@@ -42,7 +42,7 @@ def create_content_cache() -> str:
     ]
 
     content_cache = client.caches.create(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         config=CreateCachedContentConfig(
             contents=contents,
             system_instruction=system_instruction,

genai/content_cache/contentcache_use_with_txt.py

@@ -23,7 +23,7 @@ def generate_content(cache_name: str) -> str:
     # Use content cache to generate text response
     # E.g cache_name = 'projects/111111111111/locations/us-central1/cachedContents/1111111111111111111'
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="Summarize the pdfs",
         config=GenerateContentConfig(
             cached_content=cache_name,

genai/content_cache/requirements.txt

@@ -1 +1 @@
-google-genai==1.7.0
+google-genai==1.16.1

genai/controlled_generation/ctrlgen_with_class_schema.py

@@ -26,7 +26,7 @@ class Recipe(BaseModel):
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="List a few popular cookie recipes.",
         config=GenerateContentConfig(
             response_mime_type="application/json",

genai/controlled_generation/ctrlgen_with_enum_class_schema.py

@@ -29,7 +29,7 @@ class InstrumentClass(enum.Enum):
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="What type of instrument is a guitar?",
         config={
             "response_mime_type": "text/x.enum",

genai/controlled_generation/ctrlgen_with_enum_schema.py

@@ -20,7 +20,7 @@ def generate_content() -> str:
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="What type of instrument is an oboe?",
         config=GenerateContentConfig(
             response_mime_type="text/x.enum",

genai/controlled_generation/ctrlgen_with_nested_class_schema.py

@@ -36,7 +36,7 @@ class Recipe(BaseModel):
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="List about 10 home-baked cookies and give them grades based on tastiness.",
         config=GenerateContentConfig(
             response_mime_type="application/json",

genai/controlled_generation/ctrlgen_with_nullable_schema.py

@@ -51,7 +51,7 @@ def generate_content() -> str:
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents=prompt,
         config=GenerateContentConfig(
             response_mime_type="application/json",

genai/controlled_generation/ctrlgen_with_resp_schema.py

@@ -36,7 +36,7 @@ def generate_content() -> str:
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents=prompt,
         config={
             "response_mime_type": "application/json",

genai/controlled_generation/requirements.txt

@@ -1 +1 @@
-google-genai==1.7.0
+google-genai==1.16.1

genai/controlled_generation/test_controlled_generation_examples.py

@@ -26,7 +26,7 @@
 import ctrlgen_with_resp_schema
 
 os.environ["GOOGLE_GENAI_USE_VERTEXAI"] = "True"
-os.environ["GOOGLE_CLOUD_LOCATION"] = "us-central1"
+os.environ["GOOGLE_CLOUD_LOCATION"] = "global"  # "us-central1"
 # The project name is included in the CICD pipeline
 # os.environ['GOOGLE_CLOUD_PROJECT'] = "add-your-project-name"
 

genai/count_tokens/counttoken_compute_with_txt.py

@@ -20,7 +20,7 @@ def compute_tokens_example() -> int:
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     response = client.models.compute_tokens(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="What's the longest word in the English language?",
     )
 

genai/count_tokens/counttoken_resp_with_txt.py

@@ -24,7 +24,7 @@ def count_tokens_example() -> int:
 
     # Send text to Gemini
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001", contents=prompt
+        model="gemini-2.5-flash-preview-05-20", contents=prompt
     )
 
     # Prompt and response tokens count

genai/count_tokens/counttoken_with_txt.py

@@ -20,7 +20,7 @@ def count_tokens() -> int:
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     response = client.models.count_tokens(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="What's the highest mountain in Africa?",
     )
     print(response)

genai/count_tokens/counttoken_with_txt_vid.py

@@ -29,7 +29,7 @@ def count_tokens() -> int:
     ]
 
     response = client.models.count_tokens(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents=contents,
     )
     print(response)

genai/count_tokens/requirements.txt

@@ -1 +1 @@
-google-genai==1.7.0
+google-genai==1.16.1

genai/count_tokens/test_count_tokens_examples.py

@@ -24,7 +24,7 @@
 import counttoken_with_txt_vid
 
 os.environ["GOOGLE_GENAI_USE_VERTEXAI"] = "True"
-os.environ["GOOGLE_CLOUD_LOCATION"] = "us-central1"
+os.environ["GOOGLE_CLOUD_LOCATION"] = "global"  # "us-central1"
 # The project name is included in the CICD pipeline
 # os.environ['GOOGLE_CLOUD_PROJECT'] = "add-your-project-name"
 

genai/embeddings/requirements.txt

@@ -1 +1 @@
-google-genai==1.7.0
+google-genai==1.16.1

genai/express_mode/api_key_example.py

@@ -23,7 +23,7 @@ def generate_content() -> str:
     client = genai.Client(vertexai=True, api_key=API_KEY)
 
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="Explain bubble sort to me.",
     )
 

genai/express_mode/requirements.txt

@@ -1 +1 @@
-google-genai==1.7.0
+google-genai==1.16.1

genai/express_mode/test_express_mode_examples.py

@@ -40,7 +40,7 @@ def test_api_key_example(mock_genai_client: MagicMock) -> None:
 
     mock_genai_client.assert_called_once_with(vertexai=True, api_key="YOUR_API_KEY")
     mock_genai_client.return_value.models.generate_content.assert_called_once_with(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="Explain bubble sort to me.",
     )
     assert response == "This is a mocked bubble sort explanation."

genai/image_generation/requirements.txt

@@ -1,2 +1,2 @@
-google-genai==1.11.0
+google-genai==1.16.1
 pillow==11.1.0

genai/live/requirements.txt

@@ -1 +1 @@
-google-genai==1.10.0
+google-genai==1.16.1

genai/provisioned_throughput/provisionedthroughput_with_txt.py

@@ -31,7 +31,7 @@ def generate_content() -> str:
         )
     )
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="How does AI work?",
     )
     print(response.text)

genai/provisioned_throughput/requirements.txt

@@ -1 +1 @@
-google-genai==1.7.0
+google-genai==1.16.1

genai/safety/requirements.txt

@@ -1 +1 @@
-google-genai==1.7.0
+google-genai==1.16.1

genai/safety/safety_with_txt.py

@@ -54,7 +54,7 @@ def generate_content() -> GenerateContentResponse:
     ]
 
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents=prompt,
         config=GenerateContentConfig(
             system_instruction=system_instruction,

genai/safety/test_safety_examples.py

@@ -22,7 +22,7 @@
 
 
 os.environ["GOOGLE_GENAI_USE_VERTEXAI"] = "True"
-os.environ["GOOGLE_CLOUD_LOCATION"] = "us-central1"
+os.environ["GOOGLE_CLOUD_LOCATION"] = "global"  # "us-central1"
 # The project name is included in the CICD pipeline
 # os.environ['GOOGLE_CLOUD_PROJECT'] = "add-your-project-name"
 

genai/template_folder/requirements.txt

@@ -1 +1 @@
-google-genai==1.7.0
+google-genai==1.16.1

genai/template_folder/templatefolder_with_txt.py

@@ -1,4 +1,4 @@
-# Copyright 2024 Google LLC
+# Copyright 2025 Google LLC
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.

genai/template_folder/test_templatefolder_examples.py

@@ -1,4 +1,4 @@
-# Copyright 2024 Google LLC
+# Copyright 2025 Google LLC
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -16,7 +16,7 @@
 import templatefolder_with_txt
 
 os.environ["GOOGLE_GENAI_USE_VERTEXAI"] = "True"
-os.environ["GOOGLE_CLOUD_LOCATION"] = "us-central1"
+os.environ["GOOGLE_CLOUD_LOCATION"] = "global"  # "us-central1"
 # The project name is included in the CICD pipeline
 # os.environ['GOOGLE_CLOUD_PROJECT'] = "add-your-project-name"
 

genai/text_generation/requirements.txt

@@ -1 +1 @@
-google-genai==1.12.1
+google-genai==1.16.1

genai/text_generation/test_text_generation_examples.py

@@ -39,7 +39,7 @@
 import thinking_textgen_with_txt
 
 os.environ["GOOGLE_GENAI_USE_VERTEXAI"] = "True"
-os.environ["GOOGLE_CLOUD_LOCATION"] = "us-central1"
+os.environ["GOOGLE_CLOUD_LOCATION"] = "global"  # "us-central1"
 # The project name is included in the CICD pipeline
 # os.environ['GOOGLE_CLOUD_PROJECT'] = "add-your-project-name"
 
@@ -138,5 +138,7 @@ def test_textgen_with_youtube_video() -> None:
 
 
 def test_model_optimizer_textgen_with_txt() -> None:
+    os.environ["GOOGLE_CLOUD_LOCATION"] = "us-central1"
     response = model_optimizer_textgen_with_txt.generate_content()
+    os.environ["GOOGLE_CLOUD_LOCATION"] = "global"  # "us-central1"
     assert response

genai/text_generation/textgen_async_with_txt.py

@@ -1,4 +1,4 @@
-# Copyright 2024 Google LLC
+# Copyright 2025 Google LLC
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -21,7 +21,7 @@ async def generate_content() -> str:
     from google.genai.types import GenerateContentConfig, HttpOptions
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
-    model_id = "gemini-2.0-flash-001"
+    model_id = "gemini-2.5-flash-preview-05-20"
 
     response = await client.aio.models.generate_content(
         model=model_id,

genai/text_generation/textgen_chat_stream_with_txt.py

@@ -13,25 +13,23 @@
 # limitations under the License.
 
 
-def generate_content() -> str:
+def generate_content() -> bool:
     # [START googlegenaisdk_textgen_chat_stream_with_txt]
     from google import genai
     from google.genai.types import HttpOptions
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
-    chat_session = client.chats.create(model="gemini-2.0-flash-001")
-    response_text = ""
+    chat_session = client.chats.create(model="gemini-2.5-flash-preview-05-20")
 
     for chunk in chat_session.send_message_stream("Why is the sky blue?"):
         print(chunk.text, end="")
-        response_text += chunk.text
     # Example response:
     # The
     #  sky appears blue due to a phenomenon called **Rayleigh scattering**. Here's
     #  a breakdown of why:
     # ...
     # [END googlegenaisdk_textgen_chat_stream_with_txt]
-    return response_text
+    return True
 
 
 if __name__ == "__main__":

genai/text_generation/textgen_chat_with_txt.py

@@ -20,7 +20,7 @@ def generate_content() -> str:
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     chat_session = client.chats.create(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         history=[
             UserContent(parts=[Part(text="Hello")]),
             ModelContent(

genai/text_generation/textgen_config_with_txt.py

@@ -20,9 +20,10 @@ def generate_content() -> str:
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="Why is the sky blue?",
-        # See the documentation: https://googleapis.github.io/python-genai/genai.html#genai.types.GenerateContentConfig
+        # See the SDK documentation at
+        # https://googleapis.github.io/python-genai/genai.html#genai.types.GenerateContentConfig
         config=GenerateContentConfig(
             temperature=0,
             candidate_count=1,

genai/text_generation/textgen_sys_instr_with_txt.py

@@ -20,7 +20,7 @@ def generate_content() -> str:
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents="Why is the sky blue?",
         config=GenerateContentConfig(
             system_instruction=[

genai/text_generation/textgen_transcript_with_gcs_audio.py

@@ -24,7 +24,7 @@ def generate_content() -> str:
     Use speaker A, speaker B, etc. to identify speakers.
     """
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents=[
             prompt,
             Part.from_uri(

genai/text_generation/textgen_with_gcs_audio.py

@@ -23,7 +23,7 @@ def generate_content() -> str:
     Provide a concise summary of the main points in the audio file.
     """
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents=[
             prompt,
             Part.from_uri(

genai/text_generation/textgen_with_local_video.py

@@ -1,4 +1,4 @@
-# Copyright 2024 Google LLC
+# Copyright 2025 Google LLC
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -19,7 +19,7 @@ def generate_content() -> str:
     from google.genai.types import HttpOptions, Part
 
     client = genai.Client(http_options=HttpOptions(api_version="v1"))
-    model_id = "gemini-2.0-flash-001"
+    model_id = "gemini-2.5-flash-preview-05-20"
 
     # Read local video file content
     with open("test_data/describe_video_content.mp4", "rb") as fp:
@@ -29,6 +29,7 @@ def generate_content() -> str:
     response = client.models.generate_content(
         model=model_id,
         contents=[
+            Part.from_text(text="hello-world"),
             Part.from_bytes(data=video_content, mime_type="video/mp4"),
             "Write a short and engaging blog post based on this video.",
         ],

genai/text_generation/textgen_with_multi_img.py

@@ -28,7 +28,7 @@ def generate_content() -> str:
         local_file_img_bytes = f.read()
 
     response = client.models.generate_content(
-        model="gemini-2.0-flash-001",
+        model="gemini-2.5-flash-preview-05-20",
         contents=[
             "Generate a list of all the objects contained in both images.",
             Part.from_uri(file_uri=gcs_file_img_path, mime_type="image/jpeg"),