Google CEO Sundar Pichai recently congratulated the Meta team behind that company’s latest large language model, known as Llama 4 Scout and Llama 4 Maverick.
Posting on X, Pichai said, “Never a dull day in the AI world! Congrats to the Llama 4 team, Onwards!”
Llama 4 details
Llama 4 is a multimodal AI system capable of processing and integrating diverse data types, including text, video, images, and audio, while seamlessly converting content across these formats. Meta describes Llama 4 Scout and Llama 4 Maverick as its “most advanced models yet,” claiming they outperform all previous Llama generations.
Meta AI, the company’s AI-powered assistant integrated into apps such as WhatsApp, Messenger, and Instagram, has been updated to utilise Llama 4 in 40 countries.
The models are open-source and immediately available, while Llama 4 Behemoth, a “teacher model” designed to enhance the new models, remains in training. Meta asserts that Behemoth can surpass GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on STEM-focused benchmarks such as MATH-500 and GPQA Diamond.
Also Read: ETtech Explainer: How Meta's Llama 4 stacks up against Chinese AI models Qwen, DeepSeek, and Manus AI
Additionally, Meta announced an exclusive event, ‘LlamaCon,’ scheduled for April 29, where it will outline the possibilities and potential of Llama in greater detail.
Posting on X, Pichai said, “Never a dull day in the AI world! Congrats to the Llama 4 team, Onwards!”
Never a dull day in the AI world! Congrats to the Llama 4 team, Onwards!
— Sundar Pichai (@sundarpichai) April 5, 2025
Llama 4 details
Llama 4 is a multimodal AI system capable of processing and integrating diverse data types, including text, video, images, and audio, while seamlessly converting content across these formats. Meta describes Llama 4 Scout and Llama 4 Maverick as its “most advanced models yet,” claiming they outperform all previous Llama generations.
Meta AI, the company’s AI-powered assistant integrated into apps such as WhatsApp, Messenger, and Instagram, has been updated to utilise Llama 4 in 40 countries.
The models are open-source and immediately available, while Llama 4 Behemoth, a “teacher model” designed to enhance the new models, remains in training. Meta asserts that Behemoth can surpass GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on STEM-focused benchmarks such as MATH-500 and GPQA Diamond.
Also Read: ETtech Explainer: How Meta's Llama 4 stacks up against Chinese AI models Qwen, DeepSeek, and Manus AI
Additionally, Meta announced an exclusive event, ‘LlamaCon,’ scheduled for April 29, where it will outline the possibilities and potential of Llama in greater detail.
You may also like
Tiger Woods ordered Rolex for Masters accomplice before making luxury yacht gesture
Man murders wife at southwest Delhi home, stages suicide
Agartala: Tension in Kailashahar over protest rally demanding withdrawal of Waqf law
Lesser-known fruit lowers blood pressure, chronic inflammation and ageing issues
FEMA staff face lie detector tests amid DHS leak probe, fears of agency overhaul
