Meta Platforms on Saturday rolled out its latest generation of large language models, Llama 4 Scout and Llama 4 Maverick, marking a big leap in the company’s pursuit of cutting-edge AI technology.
Labeled by Meta as its “most advanced models yet” and “the best in their class for multimodality,” Llama 4 Scout and Llama 4 Maverick are designed to process and integrate text, images, video, and audio, making them versatile tools when we talk about AI.
Both models will be available as open-source software, Meta confirmed. The tech giant also offered a preview of Llama 4 Behemoth, a powerhouse system it described as “one of the smartest LLMs in the world,” intended to guide future models through advanced training processes.
Llama 4 release just arrived right on time, as competition is intense and more investors are investing in AI infrastructure. Meta is projected to spend up to $65 billion this year to bolster its AI capabilities. They’re also facing pressure from investors to demonstrate meaningful returns on their heavy investments.
The release of Llama 4 has not been without setbacks. According to The Information, Meta previously delayed the launch after the model fell short of technical benchmarks in areas like reasoning and mathematics. The company was also concerned about Llama 4’s performance compared to OpenAI’s models, particularly in generating humanlike voice interactions.
The push for stronger AI models follows the groundbreaking success of OpenAI’s ChatGPT, which reshaped the industry and triggered a wave of investment across the tech industry. As Meta continues to refine its models, the race to establish dominance in the AI space shows no signs of slowing down. But for now, ChatGPT and Gemini are on top, leading the game.
