무슨 발표인가

• 텍스트·이미지·오디오·비디오·코드를 다중모달로 처리
• Ultra·Pro·Nano 3가지 크기로 데이터센터부터 모바일까지 지원
• TPU v4·v5e로 학습

바드·픽셀8프로·검색에서 공개

원문 (영어)

Today we introduced Gemini , our largest and most capable AI model — and the next step on our journey toward making AI helpful for everyone. Built from the ground up to be multimodal , Gemini can generalize and seamlessly understand, operate across and combine different types of information, including text, images, audio, video and code.

This means it has sophisticated multimodal reasoning and advanced coding capabilities. And with three different sizes — Ultra, Pro and Nano — Gemini has the flexibility to run on everything from data centers to mobile devices. We trained Gemini at scale on our AI-optimized infrastructure using Google's Tensor Processing Units (TPUs) v4 and v5e.

Today, we also announced our most powerful and scalable TPU system to date, Cloud TPU v5p . Gemini is available in some of our core products starting today: Bard is using a fine-tuned version of Gemini Pro for more advanced reasoning, planning, understanding and more.

Pixel 8 Pro is the first smartphone engineered for Gemini Nano, using it in features like Summarize in Recorder and Smart Reply in Gboard. And we’re already starting to experiment with Gemini in Search, where it's making our Search Generative Experience (SGE) faster.

Early next year, we’ll bring Gemini Ultra to a new Bard Advanced experience. And in the coming months, Gemini will power features in more of our products and services like Ads, Chrome and Duet AI. Android developers who want to build Gemini-powered apps on-device can now sign up for an early preview of Gemini Nano, our most efficient model, via Android AICore.

Starting December 13, developers and enterprise customers will be able to access Gemini Pro via the Gemini API in Vertex AI or Google AI Studio, our free web-based developer tool.

원문: Google Blog (Technology/AI) — "Learn more about Gemini, our most capable AI model" (2024-09-27) 공식 원문: https://blog.google/innovation-and-ai/technology/ai/gemini-collection/