Llama 4 Model Lineup

Llama 4: Leading Multimodal Intelligence Newest model suite offering unrivaled speed and efficiency Llama 4 Behemoth 288B active parameters, 16 experts 2T total parameters The most intelligent teacher model for distillation Preview Llama 4 Maverick 17B active parameters, 128 experts 400B total parameters Native multimodal with 1M context length Available Llama 4 Scout 17B active parameters, 16 experts 109B total parameters Industry leading 10M context length Optimized inference Available

A promotional-style webpage displays the Llama 4 Behemoth, Maverick, and Scout language models with their parameter counts, expert counts, context lengths, and availability labels. The joke is the exaggerated presentation of an impressive new AI model lineup, making it suitable as a reaction to an unusually strong or exciting start to the year.

• @dev_meme / 6635 2025-04-05