Vertex AI Model Garden is Google Cloud's catalog for finding, customizing and deploying machine learning models in one place. The curated set of more than 200 models covers first-party options such as Imagen for text-to-image, Veo for text-to-video and image-to-video, and Chirp for speech-to-text, plus pre-trained APIs for text-to-speech, language processing, translation and vision. Open models include Gemma, CodeGemma, PaliGemma, Meta's Llama family, Mistral AI, AI21, Falcon, BERT, T5 FLAN, ViT and EfficientNet, and partner foundation models from third parties are supported as well. Models can be tuned with your own data and deployed to applications with one click.
Machine learning engineers and developers use it to compare model types and sizes before committing to one, since a larger model is not always the right choice. Enterprise AI teams get end-to-end MLOps and scaling in the same environment. Startups can apply to the Google for Startups Cloud Program for credits, and new Google Cloud customers receive up to $300 in free credits to try AI and machine learning products. Organizations with large projects can talk to Google's sales team or find a certified partner.
Typical workflows include picking an open model, fine-tuning it on company data and deploying it behind an endpoint. Product teams can add image generation, video generation or speech transcription to an application using the first-party models, while analytics groups can try translation and vision APIs. Researchers can browse specialized models such as image classifiers and language models side by side. Because everything lives within Google Cloud, deployments connect to existing infrastructure, security controls and billing.
