Unveiled last December, Gemini is finally available in Europe and in several languages, including French.
Available within Google Bard for a few weeks, Gemini initially skipped the old continent. The advanced model, in addition to only being able to converse in English, was not accessible in our regions… unless you manually modified the URL and used a VPN. But that’s all over!
What is Gemini?
Designed as a multi-modal AI model, Gemini is capable of understanding and processing a diverse range of data, from text and images to videos and audio. This versatility allows him to tackle complex tasks in fields as varied as mathematics, physics, and even computer programming.
Gemini follows the LLM trend and offers a structure in three distinct models: Ultra, Pro and Nano. Each of these models is optimized for specific applications, allowing Google to target a wide range of needs and uses. Gemini Ultra, the most powerful of the three, is intended for highly complex tasks. It demonstrated interesting performance, even outperforming human experts in some tests. Gemini Pro, meanwhile, is currently used to power Bard, Google’s chatbot, offering advanced reasoning and understanding capabilities. Finally, Gemini Nano is designed for mobile devices, providing Android users with the ability to develop applications incorporating Gemini’s capabilities.
One of the most notable features of Gemini is its ability to perform multimodal tasks natively. Unlike models like OpenAI’s GPT-4, which focus primarily on text, Gemini incorporates a more holistic understanding of data. This multimodal approach allows it to process textual, visual and auditory information simultaneously and coherently, paving the way for more integrated and interactive applications. For example, Gemini can not only understand complex text but also analyze images and generate high-quality code in various programming languages. If our impressions of the model are mixed for the moment, the fault of misleading communication from Google, it remains a relevant model due to its access to the web and its ability to verify its own answers.
How to access it?
Gemini should be accessible through Google AI Studio and Google Cloud Vertex, as well as in Google Android Studio for the Nano model. This extensive integration ensures that Gemini can be integrated into both mobile applications and complex cloud solutions.
For those who already want to access Gemini, it is possible to access the Pro version through Bard. All you have to do is go to the Bard website. To make sure you are using Gemini correctly, go to the “Latest updates” section. This should mention the recent addition of the new model and its multilingual capabilities.
Pixel 8 Pro owners can also take advantage of Gemini Nano thanks to the latest feature drop of Android 14. To be sure that you are using the latest version of Bard, go to the “Updates” section and check that the update The most recent date is December 6.
In the coming months, Google plans to integrate Gemini into many of its products and services. Currently available through Google Bard and the Google Pixel 8 Pro, Gemini will gradually roll out to other tools, such as Google Search, Google Ads and the Chrome browser. A paid version of Bard should also accommodate the Ultra version of Gemini, although Google has not given more information. Many rumors echo an imminent release of Bard Advanced, even if it is difficult to guess if the French will have access to this new model at the same time as our Shakespearean friends.