Google is expanding its Gemini models on the Vertex AI platform, making Gemini 1.0 Pro available to all Vertex AI customers. This model is designed for scaling across various AI tasks, offering a balance of quality, performance, and cost in areas like content generation, editing, summarization, and classification. Additionally, the sophisticated Gemini 1.0 Ultra, optimized for complex tasks such as complex instruction, code, reasoning, and multilingualism, is now generally available via an 'allowlist'.
Google has also introduced Gemini 1.5 Pro, a mid-sized multimodal model currently in private preview on Vertex AI. Notably, 1.5 Pro features a breakthrough experimental feature in long-context understanding, boasting the longest context window of any large-scale foundation model. This allows 1.5 Pro to process extensive information, including "1 hour of video, 11 hours of audio, codebases with over 30,000 lines of code, or over 700,000 words" in a single prompt.
Enterprises using 1.5 Pro can perform comprehensive analyses of entire code libraries in a single prompt, compare content across hours of video, and enable chatbots to hold extended conversations without losing details. This marks Google's continued efforts to enhance AI capabilities and expand their application across diverse tasks on the Vertex AI platform.