Anthropic, one of OpenAI’s biggest competitors, introduced the “Hybrid Artificial Intelligence Model”

29
Anthropic, one of OpenAI’s biggest competitors, introduced the “Hybrid Artificial Intelligence Model”

Anthropic, an artificial intelligence company, one of the biggest competitors of OpenAI, announced the new artificial intelligence model called Claude 3.7 Sonnet last night. The model is described as the company’s smartest model so far.

Claude 3.7 comes as a “logic model ğ that we often hear these days. So before response, “thinks”. The more striking part is that it is a hybrid logic model. This; It means that both real -time answers and thought about, and can give more detailed answers.

It is very ambitious in every matter, from coding to mathematics

Claude 3.7 compared to Anthropic’s statements; Mathematics, coding and finance has managed to prevent the previous models in bets. When we look at the information, OpenAI’s logic models with O1 and O3 Mini; We can see that there are similar levels in language, visual logic and problem solving. It is much more ahead of the previous 3.5. One -to -one way left Deepseek R1 in these areas.

Users will be able to activate their “reasoning” capabilities if they want to use the new model. In this way, 3.7 Sonnet will be able to give you much more detailed answers by making a short time. Unfortunately, there is also a negative side. That is the lack of real -time web search feature. It has information until October 2024. This pushes him a little behind other models.

The company also underlined that the new model focuses on facilitating the use of artificial intelligence. In many vehicles today, he was engaged in choosing different models and features. Claude 3.7 Sonnet allows you to perform every process with a single model by eliminating it.

Claude 3.7 Sonnet has opened through the Claude application for users and developers with Monday prestige. However, only priced premium subscribers will have the opportunity to achieve their logic capabilities.

The company also introduces another tool called Claude Code. This model, where a finite number of people can reach, will serve as an assistant who can write and perform tests.