Google, one of the leading companies in the artificial intelligence race, made a very important announcement yesterday. New version of the company’s major Gemini language model Gemini 2.0 officially introduced. The model comes as a new version of Gemini 1.5, which was released 10 months ago, as a competitor to OpenAI’s models such as o1.

Gemini 2.0, from Google The most advanced and capable model ever defined as. We should point out that in this model the company will take a different approach. Initially only the smaller version. Gemini 2.0 Flash future. So it doesn’t start with 2.0 Pro.

Comes with built-in text, audio and visual support

The first highlight of Gemini 2.0, which is both cheaper and faster, is now Built-in visual, audio and text support to have. Yan can now produce these three types of content. Only text support was available in Gemini 1.5 Flash. It is also mentioned that 2.0 can also use third-party applications and services.

Sound reproduction is one of the most notable features. We can say that it is similar to ChatGPT’s advanced voice mode. You can make the model speak quickly or slowly, and you can make it speak with different accents.

Of course, there are also significant performance improvements. According to the company, Gemini 2.0 Flash, Up to 2x faster on some subjects than Gemini 1.5 Pro. For example, we see that the success rate on advanced math tests is 63%. The previous model achieved 47% success here. He scored 92% on coding, 76% on general skills and 62.1% on reasoning. All before 1.5.

It will take search to the next level with artificial intelligence

Google’s main goal is of course to integrate Gemini 2.0 into its services. Chief among these is artificial intelligence-assisted search. AI overviews It’s coming. The company announced that the new model can answer much more complex questions He stated that this will take the search experience to the next level. It can help with coding problems, difficult parts of math and much more.

Gemini 2.0 Flash, starting today offered to users for free started. However, it can only be used on Gemini’s website. Obviously it should be noted that it is still experimental and therefore some errors may occur. Let’s also add that the company has started offering the model to developers.

We’re one step closer to making ‘artificial intelligence agents’ a reality

The next revolution of artificial intelligenceagents of artificial intelligence‘ is seen as. These tools, which we can also call assistants; they can reason, perform tasks and make decisions with minimal dependence on humans. In other words: they can simply do your work for you. Many companies are taking steps to produce such vehicles.

Google is one of them. Company, Gemini 2.0 has achieved an important breakthrough in this area. say. For example, Project Astra, which the company showed months ago, was made possible by this model. Astra was a visual system that could identify objects, give you directions and even tell you where you left your belongings.

The company too Project Zeeman He also announced that a new artificial intelligence agent would be called. This tool is a Chrome extension. What he does is use the browser for you. It takes control from you, you can move the cursor on the screen, press keys and fill out forms. In other words, Chrome can do everything for you. This tool is currently being tested on a limited number of people. There is no information about the release date.