Meta's finale open source AI model Llama 3.3 debuts this year: 70 billion parameters, performance comparable to 405 billion

Meta released Llama 3.3 with 70 billion parameters, though performance is comparable to Llama 3.1 with 405 billion parameters.Meta emphasized that Llama 3.3 models are more efficient and less costly, and can be run on standard workstations, reducing operational costs while delivering high-quality textual AI solutions.Llama 3.3 context length is 128K, supports multiple tool-use formats, and can be integrated with external tools and services to extend the model's functionality. (IT Home)