Artificial Intelligence Attack from Amazon: Preparing its own “reasoning” model

16
Artificial Intelligence Attack from Amazon: Preparing its own “reasoning” model
Amazon is working on a new model that will get angry with competition in the world of artificial intelligence. It turned out that the company has developed an advanced artificial intelligence model with “Reasoning” capabilities. According to Business Insider, this model is expected to be released under Nova brand until June.

Amazon’s artificial intelligence

According to sources, Amazon’s new artificial intelligence model will adopt the “hybrid reasoning” approach. In this way, the system will be able to give quick responses and realize more complex and long -term thinking processes. Recently, giants such as Google, OpenAI and Anthropic have come to the fore with similar models, and companies like Deepseek have focused on making these capabilities more efficient.

One of Amazon’s goals is to make the new model more efficient than competitors in terms of price-performance. OpenAI’s O1 model, Anthropic’s Claude 3.7 Sonnet, and Google’s Gemini 2.0 Flash Thinking model Nova model, Amazon’s claim will be much more cost -effective than the claim. The new model is expected to enter the top 5 in external criteria such as SWE, Berkeley Function Calling Leaderboard and AIME.

In the meantime; Amazon has invested more than $ 8 billion in Anthropic, which has developed Claude 3.7 Sonnet, the world’s first hybrid reasoning model. The two companies have cooperated in areas such as artificial intelligence chips and cloud computing. Amazon will be the direct rival of its important partner with its own model.