Google has changed its strategy in the field of large AI models and released the open source large model Gemna, targeting OpenAI and Meta. Gemna has models with two weight scales, 2B and 7B, and shares technology with the closed-source Gemini, marking Google's opening of a two-front combat model of closed source and open source. This move not only intensified competition with OpenAI and Meta, but also reshaped the competitive landscape of the large model market, forming a tripartite situation between Google Gemna, Meta LLama and Mistral. Gemna's open source strategy makes it freely available worldwide and shows performance advantages over Llama, further intensifying the competition between Google and Meta.
Google broke through last year's closed source strategy and released the open source large model Gemna. Gemna uses the same technology as Gemini, and has two models with weight scales of 2B and 7B respectively. Google has formed a two-front battle between closed source and open source in the field of large models, fiercely competing with OpenAI and Meta. The large model market has formed a competitive landscape between Google Gemna, Meta LLama and Mistral. Globally, Gemna is open for use and its performance exceeds Llama. At the same time, the competition between Google and Meta has gradually escalated, forming a chain of suppression, in which both open source and closed source models are in the chain of competition.
Google's move to open source Gemna not only enhances its competitiveness in the field of large models, but also provides more choices for global developers and promotes the progress and development of artificial intelligence technology. In the future, competition in the field of large models will become more intense, and the rise of open source models will further change the landscape of this field.