<ul><li>Meta releases Llama 3.3, a cost-effective and efficient AI model.</li><li>Llama 3.3 is up to the mark of previous models in qualitative measurements but provides more cost efficiency.</li><li>Llama 3.3 outperforms models like Gemini 1.5 Pro and GPT-4o in MMLU tests, increasing knowledge and math capability.</li><li>Meta invests over $10 billion in establishing an AI data center to support Llama 4 and new generation models.</li></ul>

Unexpectedly Drops LLAMA 3.3 — A 70B Model At 25X Lower Cost Than OpenAI’s GPT — 4o

Discover more