Google models on Vertex Artificial Intelligence

A concise guide to Google generative Artificial Intelligence models on Vertex Artificial Intelligence, outlining featured Gemini releases, Gemma open models, image and video models, embeddings, and MedLM variants.

This page catalogs Google generative Artificial Intelligence models available on Vertex Artificial Intelligence and in Model Garden. Featured Gemini releases include 3 Pro, 2.5 Pro, and 2.5 Flash. Gemini 3 Pro is listed as preview and is described as designed for comprehensive multimodal understanding and complex problem solving, featuring a 1 million token context window and capabilities for agentic workflows and autonomous coding. Gemini 2.5 Pro and Gemini 2.5 Flash are positioned for complex reasoning and cost-performance balance respectively, with 2.5 Flash offering Live API support for some endpoints and tools that provide insight into the model’s thinking process.

The document lists generally available and preview Gemini models, including Gemini 2.5 Pro, Gemini 2.5 Flash, Gemini 2.5 Flash Image, Gemini 2.5 Flash-Lite, Gemini 2.0 Flash, Gemini 2.0 Flash-Lite, and preview models such as Gemini 3 Pro Image and Gemini 2.5 Flash Live API. It also describes the Gemma family as open models. Gemma entries include Gemma 3n, Gemma 3, Gemma 2, Gemma, ShieldGemma 2, PaliGemma, CodeGemma, TxGemma, MedGemma, MedSigLIP, and T5Gemma, with details such as Gemma 3n supporting multimodal input (text, image, video, and audio) and text output in over 140 languages, and Gemma 3 offering a 128K context window and multilingual support.

Additional families covered include embeddings models (Embeddings for Text and Multimodal Embeddings), Imagen image generation models (Imagen 4 for Generation, Imagen 4 for Fast Generation, Imagen 4 for Ultra Generation, and Imagen 3 variants), Veo video generation models (Veo 2 Generate, Veo 3 Generate, Veo 3 Fast, Veo 3.1 Generate, and previews), and MedLM models (MedLM-medium and MedLM-large-large) noted as HIPAA-compliant options for medical question answering and summarization. The page points readers to Model Garden in the Google Cloud console to explore, test, customize, and deploy models, and references model versions and lifecycle information and pricing resources on Vertex Artificial Intelligence. Last updated 2025-11-26 UTC.

68

Impact Score

Samsung starts sampling 3 GB GDDR7 running at 36 Gbps

Samsung has begun sampling its fastest-ever GDDR7 memory at 36 Gbps in 24 Gb dies that translate to 3 GB per chip, and it is also mass producing 28.0 Gbps 3 GB modules reportedly aimed at a mid-cycle NVIDIA refresh.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.