New Vertex AI I/O models Gemini and Gemma unveiled

by

in

1. Vertex AI is Google Cloud’s fully-managed platform for leveraging models at scale and building AI agents, with a selection of over 150 first-party, open, and third-party foundation models.
2. Customers like ADT, IHG Hotels & Resorts, and ING Bank are using Vertex AI to innovate faster and deploy AI apps and agents.
3. New updates to Vertex AI announced at Google I/O ’24 include the availability of Gemini 1.5 Flash and PaliGemma models, with Imagen 3 and Gemma 2 models coming soon, along with new capabilities like context caching, controlled generation, and batch API for optimizing model performance.

Vertex AI is a fully-managed development platform by Google Cloud that offers over 150 first-party, open, and third-party foundation models for building AI agents. Customers like ADT, IHG Hotels & Resorts, and Verizon are using Vertex AI to innovate faster and deploy AI apps and agents more efficiently.

At Google I/O ‘24, several new models were announced for Vertex AI, including Gemini 1.5 Flash and PaliGemma, which are available to Cloud customers today. Imagen 3 and Gemma 2 models are also set to be released soon, offering advanced text-to-image generation and open models for various AI developer use cases.

New capabilities such as context caching, controlled generation, and batch API have been introduced to help customers optimize model performance and accelerate AI deployments in production. Agent Builder, an open-source integration, allows developers to build and deploy gen AI experiences using tools like LangChain, Firebase Genkit, and LlamaIndex on Vertex AI.

Grounding with Google Search is now generally available, allowing customers to access fresh, high-quality information to improve the accuracy of responses. By utilizing Gemini models with Google Search, customers can enhance their AI agents with the latest foundation models and information sources. Overall, Vertex AI continues to provide cutting-edge tools and capabilities to help developers create intelligent and informative AI agents efficiently.

Source link