Models / Llama-3_1-Nemotron-Ultra

Llama-3_1-Nemotron-Ultra by nvidia

GPU memory needed (smallest version of each): Llama-3_1-Nemotron-Ultra 253B about 662 GB. The cheapest runs on 8x RTX PRO 6000 from $17.63/h right now.

1 model (253B) · ↓ 1.6K downloads a month · from $17.63/h on 8x RTX PRO 6000 right now. Each model's versions do the same job; quantized ones (AWQ, FP8, 4-bit) need less GPU memory and run on cheaper machines.

Model details come from Hugging Face and are refreshed daily; prices follow the live offers. GPU memory is GridPup's estimate: the weights, room for an 8,192-token conversation (chat models) and the runtime's own memory. Large models are split across the GPUs of one machine ("4x RTX 4090").