Python: Fix Gemini client support for Gemini API and Vertex AI (#5258)

* Add Gemini and Vertex AI client support

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Address Gemini PR review feedback

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* removed sample run readme part

---------

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Evan Mattson <35585003+moonbox3@users.noreply.github.com>

python/packages/gemini/samples/README.md

@@ -14,5 +14,7 @@ This folder contains examples demonstrating how to use Google Gemini models with
 
 ## Environment Variables
 
-- `GEMINI_API_KEY`: Your Google AI Studio API key (get one from [Google AI Studio](https://aistudio.google.com/apikey))
-- `GEMINI_MODEL`: The Gemini model to use (e.g., `gemini-2.5-flash`, `gemini-2.5-pro`)
+- `GOOGLE_MODEL` or `GEMINI_MODEL`: The Gemini model to use (for example,
+  `gemini-2.5-flash-lite` or `gemini-2.5-pro`)
+- For Gemini Developer API: `GEMINI_API_KEY` or `GOOGLE_API_KEY`
+- For Vertex AI: `GOOGLE_GENAI_USE_VERTEXAI=true`, `GOOGLE_CLOUD_PROJECT`, and `GOOGLE_CLOUD_LOCATION`

python/packages/gemini/samples/__init__.py

@@ -0,0 +1 @@
+# Copyright (c) Microsoft. All rights reserved.

python/packages/gemini/samples/gemini_advanced.py

@@ -4,9 +4,9 @@
 
 Allows the model to reason through complex problems before responding.
 
-Requires the following environment variables to be set:
-- GEMINI_API_KEY
-- GEMINI_MODEL
+Requires ``GOOGLE_MODEL`` or ``GEMINI_MODEL`` and either Gemini Developer API credentials
+(``GEMINI_API_KEY`` or ``GOOGLE_API_KEY``) or Vertex AI settings
+(``GOOGLE_GENAI_USE_VERTEXAI``, ``GOOGLE_CLOUD_PROJECT``, and ``GOOGLE_CLOUD_LOCATION``).
 """
 
 import asyncio
@@ -23,10 +23,12 @@ async def main() -> None:
     """Example of extended thinking with a Python version comparison question."""
     print("=== Extended thinking ===")
 
+    # 1. Configure Gemini extended thinking for a reasoning-heavy request.
     options: GeminiChatOptions = {
         "thinking_config": ThinkingConfig(thinking_budget=2048),
     }
 
+    # 2. Create the agent with the Gemini chat client and default thinking options.
     agent = Agent(
         client=GeminiChatClient(),
         name="PythonAgent",
@@ -34,6 +36,7 @@ async def main() -> None:
         default_options=options,
     )
 
+    # 3. Stream the answer so you can see the final response as it arrives.
     query = "What new language features were introduced in Python between 3.10 and 3.14?"
     print(f"User: {query}")
     print("Agent: ", end="", flush=True)
@@ -45,3 +48,12 @@ async def main() -> None:
 
 if __name__ == "__main__":
     asyncio.run(main())
+
+"""
+Sample output:
+=== Extended thinking ===
+User: What new language features were introduced in Python between 3.10 and 3.14?
+Agent: Python 3.11 introduced exception groups and TaskGroup.
+Python 3.12 added PEP 695 type parameter syntax.
+Python 3.13-3.14 continued improving typing, performance, and developer ergonomics.
+"""

python/packages/gemini/samples/gemini_basic.py

@@ -4,9 +4,9 @@
 
 Covers both non-streaming and streaming responses.
 
-Requires the following environment variables to be set:
-- GEMINI_API_KEY
-- GEMINI_MODEL
+Requires ``GOOGLE_MODEL`` or ``GEMINI_MODEL`` and either Gemini Developer API credentials
+(``GEMINI_API_KEY`` or ``GOOGLE_API_KEY``) or Vertex AI settings
+(``GOOGLE_GENAI_USE_VERTEXAI``, ``GOOGLE_CLOUD_PROJECT``, and ``GOOGLE_CLOUD_LOCATION``).
 """
 
 import asyncio
@@ -35,6 +35,7 @@ async def non_streaming_example() -> None:
     """Runs the agent and waits for the complete response before printing it."""
     print("=== Non-streaming ===")
 
+    # 1. Create the agent with the Gemini chat client and local weather tool.
     agent = Agent(
         client=GeminiChatClient(),
         name="WeatherAgent",
@@ -42,6 +43,7 @@ async def non_streaming_example() -> None:
         tools=[get_weather],
     )
 
+    # 2. Ask the agent for a single weather lookup and print the final response.
     query = "What's the weather like in Karlsruhe, Germany?"
     print(f"User: {query}")
     result = await agent.run(query)
@@ -52,6 +54,7 @@ async def streaming_example() -> None:
     """Runs the agent and prints each chunk as it is received."""
     print("=== Streaming ===")
 
+    # 1. Create the same agent configuration for a streaming tool-call example.
     agent = Agent(
         client=GeminiChatClient(),
         name="WeatherAgent",
@@ -59,6 +62,7 @@ async def streaming_example() -> None:
         tools=[get_weather],
     )
 
+    # 2. Ask a multi-location question and stream the model output as it arrives.
     query = "What's the weather like in Portland and in Paris?"
     print(f"User: {query}")
     print("Agent: ", end="", flush=True)
@@ -76,3 +80,14 @@ async def main() -> None:
 
 if __name__ == "__main__":
     asyncio.run(main())
+
+"""
+Sample output:
+=== Non-streaming ===
+User: What's the weather like in Karlsruhe, Germany?
+Result: The weather in Karlsruhe, Germany is currently sunny with a high of 16°C.
+
+=== Streaming ===
+User: What's the weather like in Portland and in Paris?
+Agent: In Portland, it is currently rainy with a high of 11°C. In Paris, it is cloudy with a high of 27°C.
+"""

python/packages/gemini/samples/gemini_with_code_execution.py

@@ -4,9 +4,9 @@
 
 Allows the model to write and run code in a sandboxed environment to answer questions.
 
-Requires the following environment variables to be set:
-- GEMINI_API_KEY
-- GEMINI_MODEL
+Requires ``GOOGLE_MODEL`` or ``GEMINI_MODEL`` and either Gemini Developer API credentials
+(``GEMINI_API_KEY`` or ``GOOGLE_API_KEY``) or Vertex AI settings
+(``GOOGLE_GENAI_USE_VERTEXAI``, ``GOOGLE_CLOUD_PROJECT``, and ``GOOGLE_CLOUD_LOCATION``).
 """
 
 import asyncio
@@ -23,6 +23,7 @@ async def main() -> None:
     """Run the code execution example."""
     print("=== Code execution ===")
 
+    # 1. Create the agent with Gemini and the built-in code execution tool.
     agent = Agent(
         client=GeminiChatClient(),
         name="CodeAgent",
@@ -30,6 +31,7 @@ async def main() -> None:
         tools=[GeminiChatClient.get_code_interpreter_tool()],
     )
 
+    # 2. Ask for a computed answer and stream the generated code and final result.
     query = "What are the first 20 prime numbers? Compute them in code."
     print(f"User: {query}")
     print("Agent: ", end="", flush=True)
@@ -41,3 +43,10 @@ async def main() -> None:
 
 if __name__ == "__main__":
     asyncio.run(main())
+
+"""
+Sample output:
+=== Code execution ===
+User: What are the first 20 prime numbers? Compute them in code.
+Agent: The first 20 prime numbers are 2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, and 71.
+"""

python/packages/gemini/samples/gemini_with_google_maps.py

@@ -4,9 +4,9 @@
 
 Allows Gemini to retrieve location and mapping information before responding.
 
-Requires the following environment variables to be set:
-- GEMINI_API_KEY
-- GEMINI_MODEL
+Requires ``GOOGLE_MODEL`` or ``GEMINI_MODEL`` and either Gemini Developer API credentials
+(``GEMINI_API_KEY`` or ``GOOGLE_API_KEY``) or Vertex AI settings
+(``GOOGLE_GENAI_USE_VERTEXAI``, ``GOOGLE_CLOUD_PROJECT``, and ``GOOGLE_CLOUD_LOCATION``).
 """
 
 import asyncio
@@ -23,6 +23,7 @@ async def main() -> None:
     """Run the Google Maps grounding example."""
     print("=== Google Maps grounding ===")
 
+    # 1. Create the agent with Gemini and the built-in Google Maps grounding tool.
     agent = Agent(
         client=GeminiChatClient(),
         name="MapsAgent",
@@ -30,6 +31,7 @@ async def main() -> None:
         tools=[GeminiChatClient.get_maps_grounding_tool()],
     )
 
+    # 2. Ask a location-aware question and stream the grounded answer.
     query = "What are some highly rated restaurants in the city center of Karlsruhe, Germany?"
     print(f"User: {query}")
     print("Agent: ", end="", flush=True)
@@ -41,3 +43,11 @@ async def main() -> None:
 
 if __name__ == "__main__":
     asyncio.run(main())
+
+"""
+Sample output:
+=== Google Maps grounding ===
+User: What are some highly rated restaurants in the city center of Karlsruhe, Germany?
+Agent: Here are several highly rated restaurants near Karlsruhe city center,
+along with their cuisine styles and approximate walking distance.
+"""

python/packages/gemini/samples/gemini_with_google_search.py

@@ -4,9 +4,9 @@
 
 Allows Gemini to retrieve up-to-date information from the web before responding.
 
-Requires the following environment variables to be set:
-- GEMINI_API_KEY
-- GEMINI_MODEL
+Requires ``GOOGLE_MODEL`` or ``GEMINI_MODEL`` and either Gemini Developer API credentials
+(``GEMINI_API_KEY`` or ``GOOGLE_API_KEY``) or Vertex AI settings
+(``GOOGLE_GENAI_USE_VERTEXAI``, ``GOOGLE_CLOUD_PROJECT``, and ``GOOGLE_CLOUD_LOCATION``).
 """
 
 import asyncio
@@ -23,6 +23,7 @@ async def main() -> None:
     """Run the Google Search grounding example."""
     print("=== Google Search grounding ===")
 
+    # 1. Create the agent with Gemini and the built-in Google Search grounding tool.
     agent = Agent(
         client=GeminiChatClient(),
         name="SearchAgent",
@@ -30,6 +31,7 @@ async def main() -> None:
         tools=[GeminiChatClient.get_web_search_tool()],
     )
 
+    # 2. Ask a current-events style question and stream the grounded answer.
     query = "What is the latest stable release of the .NET SDK?"
     print(f"User: {query}")
     print("Agent: ", end="", flush=True)
@@ -41,3 +43,10 @@ async def main() -> None:
 
 if __name__ == "__main__":
     asyncio.run(main())
+
+"""
+Sample output:
+=== Google Search grounding ===
+User: What is the latest stable release of the .NET SDK?
+Agent: As of April 14, 2026, the latest stable release of the .NET SDK is .NET 10.0 (SDK 10.0.201).
+"""