Amazon is developing its own artificial intelligence model with advanced “reasoning” capabilities, according to Business Insider, which cites a source involved in the project.
Aiming for Hybrid Reasoning and Cost Efficiency
Initially planned for release in June under the Nova brand, this model is part of Amazon’s generative AI lineup introduced late last year. The company aims to implement a “hybrid reasoning” approach, enabling both quick responses and more complex extended thinking within a single system. Amazon has not officially commented on the project.
One of Amazon’s key priorities is cost efficiency, aiming to make the Nova reasoning model more affordable compared to competitors such as OpenAI’s o1, Anthropic’s Claude 3.7 Sonnet, and Google’s Gemini 2.0 Flash Thinking. Another goal is to place the model among the top five in external performance rankings, including SWE, the Berkeley Function Calling Leaderboard, and AIME, adds NIX Solutions.
Competing with Industry Leaders
The new AI model is being developed by Amazon’s AGI team, led by chief scientist Rohit Prasad. This move reinforces Amazon’s commitment to expanding its own AI ecosystem.
Additionally, it places Amazon in direct competition with Anthropic, despite their close partnership. Amazon has invested $8 billion in Anthropic, and the two companies continue to collaborate on areas like chip development and cloud computing.
We’ll keep you updated on Amazon’s progress as more details emerge.