Recipes · 11 configurations

Machines ready to copy

For every budget, a model, the hardware, the engine and the command to run. Memory, speed and price use the same formulas as the rest of the site: estimates, not measurements.

All11 Budget3 Enthusiast1 Compact4 Pro1 Team1 Datacenter1

Team€36,000

Team server: 70B for 10 people

Four L40S in FP8: quality almost intact and enough cache for ten parallel conversations.

Model Llama 3.3 70BHardware 4 × L40S 48 GBPrecision FP8Engine vLLMContext 8k × 10 users

Memory99 / 192 GB

Estimated speed~20 tok/s

Power1,400 W

Memory use52 %

vllm serve meta-llama/Llama-3.3-70B-Instruct --quantization fp8 --tensor-parallel-size 4 --max-model-len 8192

Load in the configurator The model The hardware