Google has announced new AI language models under the name Gemini. Gemini 1.0, the first version of the model, has three different variants right away: Gemini Ultra, Pro and Nano.
Google said this is its most flexible model yet, which can run efficiently on everything from mobile devices to data centres.
Until now, the standard approach to creating multimodal models involved training individual components for different modalities. These were then put together to mimic a particular functionality. These models are sometimes good at performing certain tasks, such as describing images, but struggle with more conceptual and complex reasoning.
Gemini is designed to be multimodal by nature. This allows Gemini to seamlessly understand and reason about all kinds of input from the ground up.
Gemini 1.0 is now being introduced on several products and platforms: As of today, Bard with Gemini Pro is available in English in more than 170 countries. In it, tasks such as understanding and summarizing, reasoning, brainstorming, writing and planning have been greatly improved.
Gemini also comes to Pixel smartphones. The Pixel 8 Pro is the first smartphone to use Gemini Nano, which adds a number of new features, such as the ability to summarise in the Recorder app and Smart Reply in Gboard with WhatsApp,
In the coming months, Gemini will also become available in other Google products such as Search, Ads, Chrome and Duet AI.
Starting 13 December, developers and customers can access Gemini Pro through the Gemini API in Google AI Studio and Vertex AI, Google AI Studio is a free, Web-based tool for developers that allows developers to quickly prototype and launch apps.
Android developers can also build with Gemini Nano via AICore. AICore is a new system feature available in Android 14, starting with Pixel 8 Pro devices.
Subscribers 0
Fans 0
Followers 0
Followers