Meta has launched Llama 4, a competitive large language model against Google and OpenAI. This suite includes Llama 4 Scout, which manages extensive data efficiently with 17 billion active parameters, and Llama 4 Behemoth, boasting 288 billion active parameters.
Meta claims Llama 4 models outperform rivals in performance tests, including Google’s Gemma 3 and OpenAI’s ChatGPT-4o. Llama 4 Behemoth, with almost two trillion total parameters, specializes in tasks requiring extensive computational skills.
Leave a Reply