diff --git a/gemini/use-cases/retrieval-augmented-generation/utils/intro_multimodal_rag_utils.py b/gemini/use-cases/retrieval-augmented-generation/utils/intro_multimodal_rag_utils.py
index df5873dc29e..d9183c4ebfc 100644
--- a/gemini/use-cases/retrieval-augmented-generation/utils/intro_multimodal_rag_utils.py
+++ b/gemini/use-cases/retrieval-augmented-generation/utils/intro_multimodal_rag_utils.py
@@ -20,7 +20,7 @@
 from vertexai.vision_models import Image as vision_model_Image
 from vertexai.vision_models import MultiModalEmbeddingModel
 
-text_embedding_model = TextEmbeddingModel.from_pretrained("textembedding-gecko@latest")
+text_embedding_model = TextEmbeddingModel.from_pretrained("textembedding-gecko@003")
 multimodal_embedding_model = MultiModalEmbeddingModel.from_pretrained(
     "multimodalembedding@001"
 )