Llama 3.3 70B

Llama 3.3 70B is Meta's most capable model in the Llama 3 series at the 70B parameter scale. It delivers performance competitive with much larger models like Llama 3.1 405B on many benchmarks, particularly in reasoning, coding, and multilingual tasks. This model represents a significant efficiency improvement, offering near-frontier performance in a size that can run on high-end consumer hardware with appropriate quantization. It supports 128K context and excels at instruction following, tool use, and complex reasoning.

Compatible Hardware for Q4_K_M

Showing compatibility for the recommended quantization (Q4_K_M, 43.5 GB VRAM).