Amazon is making substantial investments in the development of a large language model (LLM) known internally as "Olympus." With an impressive 2 trillion parameters, this model aims to compete with leading models from OpenAI and Alphabet. The project is led by Rohit Prasad, former head of Alexa, who now reports directly to CEO Andy Jassy. Amazon's focus on LLMs aligns with its strategy to bolster its offerings on AWS, catering to enterprise clients seeking top-performing models. The company has previously trained smaller models like Titan and partnered with AI startups like Anthropic and AI21 Labs, integrating their solutions into AWS. Despite the cost-intensive nature of training larger AI models, Amazon is prioritizing investments in LLMs and generative AI, signaling its commitment to advancing AI capabilities. Amazon declined to provide official comments, and specific details about the project remain undisclosed.