Meta’s latest open source AI models challenge GPT, Gemini, and Claude

Meta has announced the latest repetition of the AI ​​Model Open Open OpenS family, which was developed by the brand while competing in the industrial intelligence industry in its intensification.

The new AI family has four models, a detailed Meta Llama 4 Scout, Llama 4 MAVERICK and Llama 4 Behemoth. Meta is detailed on its website from artificial intelligence that the models were trained on “large quantities of texts, image and videos that are not used for them.” This indicates that the models will have a variety of multimedia capabilities.

Currently, two models- Llama 4 Scout, Llama 4 MAVERICK- are available for users to access through Meta platforms, including WhatsApp, Messenger and Instagram Direct, as well as Meta’s Ai, Llama.com, starting from Saturday. Developers can also access artificial intelligence models in open source warehouses like Huging Face. The Llama 4 Beheemoth model has not been launched and is still in training. The company indicated that the industry must expect the giant model to outperform similar models, and that it will serve as a teacher for other models in the Llama 4 family.

Amid its internal test, Meta compared Llama 4 models to artificial intelligence techniques to determine their capabilities and the best cases of use. The company indicated that Llama 4 MAVERICK will work better in creative writing. In addition, it outperformed Openai GPT-4O and Google Gemini 2.0 models in functions such as coding, logic, multi-language generation, and images. Meanwhile, Maverick has struggled to keep pace with the capabilities of artificial intelligence models, including Gemini 2.5 Pro, GPT-4.5 and Claude 3.7 Sonnet.

Although Meta claims that the giant can outperform most of these models, with the exception of Gemini 2.5 Pro, the company has not been able to reduce the costs of devices to train its strongest model.

Techcrunch pointed out that the interest gained by the Chinese company Deepseek from artificial intelligence from its unique competitive models that put Meta in the notification. The company is said to have been studying extensively how the competing company developed its prominent models, including R1 and V3, with low operating costs of previous Llama models.

The company detailed that the Llama 4 Scout model can work on the NVIDIA H100 graphics processing unit. Llama 4 MAVERICK can be run on the NVIDIA H100 DGX graphics system.

Meta is scheduled to host the first conference in Llamacon AI on April 29. The company also has a independent Chatbot Meta AI that is scheduled to be launched in the second quarter of the year, according to CNBC.

Meta is not the only company that ginger enjoys with the timeline of the main artificial intelligence models. Openai recently modified the launch of the GPT-5, with the CEO of the company, Sam Altman, on social media, fans must expect O3 and O4-MINI thinking models in the coming weeks as an alternative to GPT-5. The executive joint is that the GPT-5 will be launched now in the coming months, which will give Openai additional time to get the model to a record level.

trends">