Llama 3.3 70B Instruct · 8× NVIDIA GeForce RTX 4090 · fp16

$/hr: $2.80
Weights / VRAM: 142 / 192 GB
FP16 TFLOPS: 1,320
Params: 70.6B · llama
Context: 131,072
License: llama3.3

← all offerings · deployments