Recently, there has been a big news from the technology industry that Amazon is secretly developing an artificial intelligence model with advanced reasoning capabilities. The new product, called Nova, is expected to be officially unveiled in June this year, and it will become a key member of Amazon's series of generative AI models released late last year. According to people familiar with the matter, the Nova model will adopt innovative "hybrid reasoning" technology that can achieve the dual functions of fast response and complex reasoning in a single system.
In the field of artificial intelligence, inference models are regarded as one of the most promising cutting-edge technologies. Although such models usually require longer processing time, they can effectively solve complex problems through multi-dimensional solutions and thinking chain technologies. At present, technology giants such as Google, OpenAI and Anthropic have launched their respective inference models, and market competition is becoming increasingly fierce. Amazon hopes to gain an advantage in cost-effectiveness through the Nova model, and strives to surpass competitors such as OpenAI's o1, Anthropic's Claude 3.7 Sonnet and Google's Gemini2.0 Flash Thinking in pricing strategies.
One of Amazon's advantages in the field of artificial intelligence is the cost-effectiveness of its Nova model. It is reported that the model is 75% cheaper than the third-party model provided through the Bedrock AI platform. Additionally, Amazon has set ambitious goals to hope that the Nova model will be among the top five in multiple external benchmark evaluations. These evaluations include software development capabilities (SWE), Berkeley function call rankings, and AIME and other mathematical skills tests, fully demonstrating Amazon's technical strength in inference model development.
Amazon chief scientist Rohit Prasad said the company is continuing to increase investment in the field of artificial intelligence. Although Amazon provides a diverse model choice through the Bedrock platform, the company is still committed to independent research and development. It is worth mentioning that Amazon maintains a close partnership with Anthropic, which recently launched the Claude 3.7 Sonnet model. The two companies have conducted in-depth cooperation in the fields of artificial intelligence chips and cloud computing, with Amazon's total investment in Anthropic reaching US$8 billion.
Overall, Amazon's R&D investment in the field of inference models demonstrates its firm determination in the development of artificial intelligence technology. This strategic move will not only promote Amazon's innovation in the field of AI, but will also intensify its technological competition with its major competitors, injecting new vitality into the development of the artificial intelligence industry.