Model Garden isn't just Gemini's catalog -- it's the only place in Google Cloud where Gemini, Claude, Llama, and Imagen coexist in the same deployment interface.
Vertex AI offers more than 200 first-party, third-party, and open source models, including Gemini, Imagen, Claude, and Llama -- Model Garden isn't exclusive to Google's own models, it's the unified entry point for deploying practically any relevant model within the GCP ecosystem.
You can filter the catalog by model type and characteristics, including open models and one-click deployment. For open models specifically, there are dedicated open-model and deployment filters -- in the model collections section of the filter panel, you can select "Self-deploy partner models" to see every model you can self-deploy directly.
Once you've filtered the catalog, clicking a model's name opens its model card -- with the model name, description, supported tasks, and deployment options, giving you the information needed to make an informed decision before deploying. It's Google's equivalent of the Azure AI Foundry leaderboard: a centralized source of information before committing to a model.
Model Garden is the discovery and deployment catalog; Vertex AI Studio is the environment where you experiment with prompts and evaluate behavior once you've already chosen a model -- it's a common mistake to confuse them, but they fill sequential, distinct roles in the workflow: first you choose in Model Garden, then you iterate in Studio.
Model Garden also lets you list supported Hugging Face models, and filter them by model name -- an additional path to accessing models that aren't part of Google's curated catalog but that the open source community has already extensively validated.
Carlos Montiel is an enterprise AI solutions architect. He implements LLMs, Agents, RAG, and orchestrators for companies across Guatemala and Latin America. Reach out for a consultation.
Contact Carlos Montiel