In a notable display of camaraderie within the tech industry, Google CEO Sundar Pichai publicly congratulated Meta on the launch of its latest AI models, Llama 4, on April 6, 2025. This marks a significant moment in the rapidly evolving artificial intelligence landscape, as Meta's advancements aim to position the company as a formidable competitor against established players like Google and OpenAI.
The Llama 4 suite comprises three models: Llama 4 Maverick, Llama 4 Scout, and Llama 4 Behemoth, with the latter currently in training. The two available models, Scout and Maverick, are designed to enhance Meta's AI capabilities across its platforms, including WhatsApp, Messenger, and Instagram. Pichai's enthusiastic tweet read, "Never a dull day in the AI world! Congrats to the Llama 4 team, Onwards!" This message reflects the competitive spirit that fuels innovation in the tech sector.
Meta claims that its Llama 4 Scout model outperforms Google's own AI offerings, specifically the Gemma 3 and Gemini 2.0 Flash-Lite models, as well as the open-source Mistral 3.1. The Scout model operates on a single Nvidia H100 GPU and boasts a remarkable 10-million-token context window, allowing it to process extensive documents efficiently. With a total of 109 billion parameters and 17 billion active parameters, Scout's performance on various benchmarks has been impressive, underscoring Meta's commitment to advancing AI technology.
On the other hand, the Llama 4 Maverick model requires a more robust computing setup, such as an Nvidia H100 DGX system. Meta asserts that Maverick competes directly with OpenAI's GPT-4o and Google's Gemini 2.0 Flash in critical areas like coding, reasoning, multilingual capabilities, and image processing. Maverick features a staggering 400 billion total parameters, with 17 billion active parameters distributed across 128 "experts." This innovative "mixture of experts" (MoE) architecture enhances efficiency by activating only the relevant parts of the model for specific tasks.
Looking ahead, the Llama 4 Behemoth model, which is still under development, is expected to be a game-changer in the AI landscape. It will feature 288 billion active parameters and nearly two trillion total parameters. Meta's internal testing indicates that Behemoth outperforms several high-profile models, including GPT-4.5 and Claude 3.7 Sonnet, in various STEM evaluations. This ambitious project highlights Meta's drive to push the boundaries of AI capabilities.
In a related context, Pichai also congratulated Elon Musk on the launch of the Grok-3 AI model, expressing his eagerness to try it out. Musk, the owner of X, claimed that Grok 3 surpasses several leading AI models, including Gemini 2 Pro, DeepSeek V3, and GPT-4o, particularly in areas such as science, coding, and mathematics. Pichai's supportive comments toward Musk and Meta illustrate a collaborative environment despite the intense competition in the tech industry.
Ahmad Al-Dale, the head of generative AI at Meta, responded to Pichai's congratulatory post, saying, "Thanks Sundar! We don’t like dull days here!" This exchange not only highlights the friendly rivalry between these tech giants but also reflects the shared enthusiasm for pushing the boundaries of AI technology.
Meta's vision for its AI initiatives is ambitious. During the launch of the Llama 4 models, Meta CEO shared the company's goal to build the world's leading AI, emphasizing the importance of open-sourcing technology to ensure universal accessibility. "I’ve said for a while that I think that open source AI is going to become the leading models, and with a lot with Llama 4 this is starting to happen," he stated in a video message. This commitment to open-source AI could reshape the landscape of artificial intelligence, making advanced capabilities available to a broader audience.
As the competition heats up in the AI sector, the release of Llama 4 models signifies a pivotal moment for Meta. By leveraging advanced technology and innovative architectures, Meta aims to carve out a significant share of the AI market. The response from industry leaders like Pichai and Musk indicates a recognition of the importance of these advancements and the potential impact they may have on the future of technology.
In conclusion, the launch of Meta's Llama 4 models represents a bold step forward in the AI arena, challenging established players and setting new benchmarks for performance and efficiency. As companies like Meta, Google, and others continue to innovate, the landscape of artificial intelligence will undoubtedly evolve, leading to new opportunities and challenges for developers, businesses, and consumers alike.
