diff --git a/litellm/llms/vertex_ai/multimodal_embeddings/transformation.py b/litellm/llms/vertex_ai/multimodal_embeddings/transformation.py
index d82c2bebb7f..d5cfca96812 100644
--- a/litellm/llms/vertex_ai/multimodal_embeddings/transformation.py
+++ b/litellm/llms/vertex_ai/multimodal_embeddings/transformation.py
@@ -201,6 +201,11 @@ class VertexAIMultimodalEmbeddingConfig(BaseEmbeddingConfig):
 
             request_data["instances"] = [vertex_request_instance]
 
+        if "outputDimensionality" in optional_params:
+            request_data["parameters"] = {
+                "outputDimensionality": optional_params["outputDimensionality"]
+            }
+
         return cast(dict, request_data)
 
     def transform_embedding_response(
diff --git a/litellm/types/llms/vertex_ai.py b/litellm/types/llms/vertex_ai.py
index c49fc96a65b..e5442e692b9 100644
--- a/litellm/types/llms/vertex_ai.py
+++ b/litellm/types/llms/vertex_ai.py
@@ -514,8 +514,9 @@ class Instance(TypedDict, total=False):
     video: InstanceVideo
 
 
-class VertexMultimodalEmbeddingRequest(TypedDict):
-    instances: List[Instance]
+class VertexMultimodalEmbeddingRequest(TypedDict, total=False):
+    instances: Required[List[Instance]]
+    parameters: dict
 
 
 class VideoEmbedding(TypedDict):
diff --git a/tests/test_litellm/llms/vertex_ai/multimodal_embeddings/test_vertex_ai_multimodal_embedding_transformation.py b/tests/test_litellm/llms/vertex_ai/multimodal_embeddings/test_vertex_ai_multimodal_embedding_transformation.py
index 63677c0f5f1..e060e2a7e1a 100644
--- a/tests/test_litellm/llms/vertex_ai/multimodal_embeddings/test_vertex_ai_multimodal_embedding_transformation.py
+++ b/tests/test_litellm/llms/vertex_ai/multimodal_embeddings/test_vertex_ai_multimodal_embedding_transformation.py
@@ -135,3 +135,35 @@ class TestVertexMultimodalEmbedding:
         ]
         result = self.config.process_openai_embedding_input(input_data)
         assert result == expected_output, f"Expected {expected_output}, but got {result}"
+
+    def test_dimensions_parameter_forwarded_as_output_dimensionality(self):
+        """Test that the dimensions parameter is forwarded as outputDimensionality
+        in the request body parameters field.
+
+        Fixes: https://github.com/BerriAI/litellm/issues/24392
+        """
+        optional_params = {}
+        optional_params = self.config.map_openai_params(
+            non_default_params={"dimensions": 128},
+            optional_params=optional_params,
+            model="multimodalembedding",
+            drop_params=False,
+        )
+        request = self.config.transform_embedding_request(
+            model="multimodalembedding",
+            input="Hello world",
+            optional_params=optional_params,
+            headers={},
+        )
+        assert "parameters" in request, "Request should contain 'parameters' field"
+        assert request["parameters"]["outputDimensionality"] == 128
+
+    def test_no_parameters_field_when_dimensions_not_set(self):
+        """Test that no parameters field is added when dimensions is not set."""
+        request = self.config.transform_embedding_request(
+            model="multimodalembedding",
+            input="Hello world",
+            optional_params={},
+            headers={},
+        )
+        assert "parameters" not in request