As the world of technology expands daily, various artificial intelligence companies (AI) launch new and advanced innovations. Recently, TECH GIANT META launched the latest open linguistic models (LLM), Llama 4.
The company claims that the newly launched Lama 4 is an open -weight answer for its competing companies such as Deepseek, Google, Openai, etc.
What is the Llama 4 AI model that Meta launched?
Llama 4 is a new collection of artificial intelligence models that will make Meta Ai more intelligent on applications such as WhatsApp, Instagram and Messenger. Two models, Llama 4 Scout and Llama 4 MAVERICK, from Meta’s Llama or its face can be downloaded. They also showed Llama 4 Beheemoth, a special model that helps in training other Llama 4 models.
Simply put, the models available for the audience are known as the weights of open -weight artificial intelligence models. Moreover, users can also operate them locally without relying on cloud application programming facades. Despite its availability, models may still have licensing restrictions on modifications or commercial use.
according to Reports, Meta Llama 4 Scout, a smaller model of artificial intelligence, designed to be multimedia. A mixture of experts (MEE), which resembles a team of 16 specialized experts working together, uses 17 billion active teachers to make them smart.
It can process a large amount of information simultaneously, up to 10 million pieces of data such as words or parts of images. In addition, it is effective enough to run on the NVIDIA H100 graphics processing unit only one powerful, which is a type of computer chips designed for heavy artificial intelligence tasks.
Llama 4 can be the best mathematics teacher
Llama 4 Beheemoth is the largest and most powerful model in the Llama 4 family, behaves like a teacher to help train smaller models. It is also a multimedia, which can handle the text, images, etc.
It has 288 billion active teachers and is a total of nearly a trillion of the teacher in general. Meta claims to be the best performance in things like mathematics, work with different languages, and photo tasks, which makes them ideal for other Llama 4 models.
In addition, Meta also claimed that the new artificial intelligence models can overcome the Gemma 3 of Google, Gemini 2.0 Flash-Lite, Openai’s Chatgpt-4O and Mistral 3.1 in different tests that measure the performance of artificial intelligence models.
Moreover, Google’s CEO of Google, Sundar Pichai, also participated X post Congratulations to the Llama model.
Also read: O3 and O4-MINI are launched soon, GPT-5 to follow: Sam Altman
