Meta Launches Next-Gen Llama-4 AI Suite to Compete in Global AI Race
| Category | Details |
|---|---|
| Company | Meta, led by CEO Mark Zuckerberg |
| Event | Launch of Llama-4 (Large Language Model) suite |
| Launch Date | April 6, 2025 |
| Models Introduced | Scout (best multimodal model), Maverick (flagship AI assistant), Behemoth (teacher model) |
| Capabilities | Advanced reasoning (text, images, video), document summarization, code generation, visual processing, multimodal tasks |
| Performance Claims | Outperforms Gemma 3, Gemini 2.0, Flash Lite, Mistral 3.1; Maverick exceeds GPT-4o and Gemini 2.0; Scout hailed as most accurate multimodal AI |
| Availability | Scout and Maverick freely available; Behemoth in preview stage |
| Investment | $65 billion targeted for AI development in 2025 |
| Key Projects | $10 billion data center in Louisiana; acquisition of high-performance chips; hiring talent for AI research |
| Goal | Make Meta AI the most widely used AI assistant globally by end of 2025; integrate into social media platforms; expand to AI-powered smart glasses and wearables |
| Global Competition | Pressure from China's DeepSeek (models R1 and V3 surpassed Llama-2 in benchmarks); Meta accelerated Llama-4 development with "AI war rooms"; Behemoth designed to counter DeepSeek's efficiency and cost innovations |