[Feat] VertexAI Gemma model family streaming support + Added MedGemma (#15427)

* test_acompletion_filters_stream_and_stream_options

* fix: stream_options

* docs medgemma

* lint fix

* docs

docs/my-website/docs/providers/vertex_batch.md

@@ -1,7 +1,7 @@
 import Tabs from '@theme/Tabs';
 import TabItem from '@theme/TabItem';
 
-## **Batch APIs**
+# Vertex Batch APIs
 
 Just add the following Vertex env vars to your environment. 
 

docs/my-website/docs/providers/vertex_self_deployed.md

@@ -124,10 +124,7 @@ Deploy Gemma models on custom Vertex AI prediction endpoints with OpenAI-compati
 | Vertex Documentation | [Vertex AI Prediction](https://cloud.google.com/vertex-ai/docs/predictions/get-predictions) |
 | Required Parameter | `api_base` - Full prediction endpoint URL |
 
-### Usage
-
-<Tabs>
-<TabItem value="proxy" label="Proxy">
+**Proxy Usage:**
 
 **1. Add to config.yaml**
 
@@ -160,9 +157,7 @@ curl http://0.0.0.0:4000/v1/chat/completions \
   }'
 ```
 
-</TabItem>
-
-<TabItem value="sdk" label="SDK">
+**SDK Usage:**
 
 ```python
 from litellm import completion
@@ -176,5 +171,59 @@ response = completion(
 )
 ```
 
-</TabItem>
-</Tabs>
+## MedGemma Models (Custom Endpoints)
+
+Deploy MedGemma models on custom Vertex AI prediction endpoints with OpenAI-compatible format. MedGemma models use the same `vertex_ai/gemma/` route.
+
+| Property | Details |
+|----------|---------|
+| Provider Route | `vertex_ai/gemma/{MODEL_NAME}` |
+| Vertex Documentation | [Vertex AI Prediction](https://cloud.google.com/vertex-ai/docs/predictions/get-predictions) |
+| Required Parameter | `api_base` - Full prediction endpoint URL |
+
+**Proxy Usage:**
+
+**1. Add to config.yaml**
+
+```yaml
+model_list:
+  - model_name: medgemma-model
+    litellm_params:
+      model: vertex_ai/gemma/medgemma-2b-v1
+      api_base: https://ENDPOINT.us-central1-PROJECT.prediction.vertexai.goog/v1/projects/PROJECT_ID/locations/us-central1/endpoints/ENDPOINT_ID:predict
+      vertex_project: "my-project-id"
+      vertex_location: "us-central1"
+```
+
+**2. Start proxy**
+
+```bash
+litellm --config /path/to/config.yaml
+```
+
+**3. Test it**
+
+```bash
+curl http://0.0.0.0:4000/v1/chat/completions \
+  -H "Content-Type: application/json" \
+  -H "Authorization: Bearer sk-1234" \
+  -d '{
+    "model": "medgemma-model",
+    "messages": [{"role": "user", "content": "What are the symptoms of hypertension?"}],
+    "max_tokens": 100
+  }'
+```
+
+**SDK Usage:**
+
+```python
+from litellm import completion
+
+response = completion(
+    model="vertex_ai/gemma/medgemma-2b-v1",
+    messages=[{"role": "user", "content": "What are the symptoms of hypertension?"}],
+    api_base="https://ENDPOINT.us-central1-PROJECT.prediction.vertexai.goog/v1/projects/PROJECT_ID/locations/us-central1/endpoints/ENDPOINT_ID:predict",
+    vertex_project="my-project-id",
+    vertex_location="us-central1",
+)
+```

docs/my-website/sidebars.js

@@ -423,6 +423,7 @@ const sidebars = {
           items: [
             "providers/vertex",
             "providers/vertex_partner",
+            "providers/vertex_self_deployed",
             "providers/vertex_image",
             "providers/vertex_batch",
           ]

litellm/llms/vertex_ai/vertex_gemma_models/transformation.py

@@ -13,10 +13,9 @@ from typing import Any, Callable, Dict, List, Optional, Union, cast
 import httpx
 
 from litellm.llms.base_llm.chat.transformation import BaseLLMException
-from litellm.llms.custom_httpx.http_handler import get_async_httpx_client
 from litellm.llms.openai.chat.gpt_transformation import OpenAIGPTConfig
 from litellm.types.llms.openai import AllMessageValues
-from litellm.types.utils import LlmProviders, ModelResponse
+from litellm.types.utils import ModelResponse
 
 
 class VertexGemmaConfig(OpenAIGPTConfig):
@@ -88,6 +87,7 @@ class VertexGemmaConfig(OpenAIGPTConfig):
         # Remove params not needed/supported by Vertex Gemma
         openai_request.pop("model", None)
         openai_request.pop("stream", None)  # Streaming not supported, will be faked client-side
+        openai_request.pop("stream_options", None)  # Stream options not supported
         
         # Wrap in Vertex Gemma format
         return {

tests/test_litellm/llms/vertex_ai/vertex_gemma_models/test_vertex_gemma_transformation.py

@@ -309,3 +309,83 @@ class TestVertexGemmaCompletion:
             assert len(chunk.choices) > 0
             assert chunk.choices[0].delta.content == "Streaming test response"
 
+    @pytest.mark.asyncio
+    async def test_acompletion_filters_stream_and_stream_options(self):
+        """
+        Test that both stream and stream_options are filtered out from the request.
+        
+        Verifies that when stream=True and stream_options={'include_usage': True} are passed,
+        neither parameter is sent to the Vertex API since Vertex Gemma doesn't support them.
+        """
+        # Mock Vertex response
+        mock_vertex_response = {
+            "deployedModelId": "1207280419999999999",
+            "model": "projects/993702345710/locations/us-central1/models/gemma-3-12b-it-1222199011122",
+            "modelDisplayName": "gemma-3-12b-it-1222199011122",
+            "modelVersionId": "1",
+            "predictions": {
+                "choices": [
+                    {
+                        "finish_reason": "stop",
+                        "index": 0,
+                        "logprobs": None,
+                        "message": {
+                            "content": "Test response",
+                            "reasoning_content": None,
+                            "role": "assistant",
+                            "tool_calls": [],
+                        },
+                        "stop_reason": None,
+                    }
+                ],
+                "created": 1759863903,
+                "id": "chatcmpl-test",
+                "model": "google/gemma-3-12b-it",
+                "object": "chat.completion",
+                "prompt_logprobs": None,
+                "usage": {
+                    "completion_tokens": 2,
+                    "prompt_tokens": 10,
+                    "prompt_tokens_details": None,
+                    "total_tokens": 12,
+                },
+            },
+        }
+        
+        with patch(
+            "litellm.llms.custom_httpx.http_handler.get_async_httpx_client"
+        ) as mock_get_client:
+            mock_client = Mock()
+            mock_response = Mock()
+            mock_response.status_code = 200
+            mock_response.json.return_value = mock_vertex_response
+            mock_client.post = AsyncMock(return_value=mock_response)
+            mock_get_client.return_value = mock_client
+            
+            # Call with both stream and stream_options
+            response = await litellm.acompletion(
+                model="vertex_ai/gemma/gemma-3-12b-it-1222199011122",
+                messages=[{"role": "user", "content": "Test"}],
+                stream=True,
+                stream_options={"include_usage": True},
+                api_base="https://test.us-central1-project.prediction.vertexai.goog/v1/projects/PROJECT_ID/locations/us-central1/endpoints/ENDPOINT_ID:predict",
+                vertex_project="PROJECT_ID",
+                vertex_location="us-central1",
+            )
+            
+            # Verify the request sent to Vertex
+            call_args = mock_client.post.call_args
+            assert call_args is not None, "HTTP client was not called"
+            
+            request_data = call_args.kwargs["json"]
+            print("request body=", json.dumps(request_data, indent=4))
+            instance = request_data["instances"][0]
+            
+            # Critical: Verify both stream and stream_options are NOT sent to Vertex API
+            assert "stream" not in instance, "stream parameter should not be sent to Vertex API"
+            assert "stream_options" not in instance, "stream_options parameter should not be sent to Vertex API"
+            
+            # Verify other parameters are present
+            assert "messages" in instance
+            assert instance["@requestFormat"] == "chatCompletions"
+