Google CEO’s Sundar Pichai emphasizes the efficiency of the company’s GEMINI AI model, especially for the DeepSeek models that created the waves of the AI industry for the reported financial abilities. We emphasize the benefits of performance.
Pichai acknowledged the “impressive job” of the Deepseek team during the fourth quarter revenue call of Google-Parent Company Alphabet. I think they did a very good job, “said Pichai.
But he emphasized that Google focuses on promoting efficiency in the provision of AI models, emphasizing Gemini’s strengths that balance costs, performance and delay.
“Looking at one of the areas where the gemini model shines, it is a paratate frontier of the cost performance of Reiitenshi, and if you look at all three attributes, we will do it -we have this Frontier period. I’m leading, “said the executive.
The GEMINI 2.0 flash model is more efficient than Deepseek: pichai
Executive also pointed out the efficiency of the Gemini 2.0 Flash model and the Gemini 2.0 Flash Thinking model, and compared them with DeepSeek’s V3 and R1 models.
“And I think both 2.0 flash models, 2.0 flash thinking model, are part of the most efficient models, including V3 and R1 in DeepSeek,” he added.
“And many of them are the strength of the end -to -end of the full stack development and the strength of attachment to the cost per query. I think all are set for workloads. 。
Executive has also emphasized the increase in spending spending on inferences compared to the past three years of training.
“Some of the reasons for being very excited about the AI opportunity are actually reduced the cost of using it, making it possible to execute more use cases, so that you can drive extraordinary use cases. That’s why it’s just as big as the opportunity to meet. Was concluded.