>>

Run Llama-3.1-405B on Lium

Meta · Text generation · 405.9B parameters · 811.7 GB BF16 weights · released 2024-07-16
It needs about 975 GB of VRAM. The cheapest setup that runs it is 8x B200 (1440 GB), from $29.60 per hour, billed per second.

How to run it

Rent the pod above, open a terminal over SSH or Jupyter and serve meta-llama/Llama-3.1-405B:
vllm serve meta-llama/Llama-3.1-405B --tensor-parallel-size 8
Claude Code and Codex can do all of it through the Lium CLI.

Frequently asked questions

How much VRAM does Llama-3.1-405B need?

About 811.7 GB for the released BF16 weights, so plan on 975 GB with room for activations and KV cache at a short context. Longer contexts and bigger batches need more.

What GPU should I rent to run Llama-3.1-405B?

8x B200 (1440 GB). On Lium that setup costs from $29.60 per hour, billed per second with no minimum.
All open models
For AI agents (CLI install, agent skill, headless signup): llms-full.txt