Run granite-docling-258M on Lium

IBM · Vision and text · 0.3B parameters · 0.5 GB BF16 weights · released 2025-05-19

It needs about 1 GB of VRAM. The cheapest setup that runs it is RTX A4500 (20 GB), from $0.16 per hour, billed per second.

Rent RTX A4500

Other GPUs that run it

RTX 5060 Ti	16 GB	from $0.17 / hour	Rent
Tesla V100	16 GB	from $0.21 / hour	Rent

How to run it

Rent the pod above, open a terminal over SSH or Jupyter and serve ibm-granite/granite-docling-258M:

vllm serve ibm-granite/granite-docling-258M

Claude Code and Codex can do all of it through the Lium CLI.

Frequently asked questions

How much VRAM does granite-docling-258M need?

About 0.5 GB for the released BF16 weights, so plan on 1 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 granite-docling-258M?

RTX A4500 (20 GB). On Lium that setup costs from $0.16 per hour, billed per second with no minimum.

All open models

For AI agents (CLI install, agent skill, headless signup): llms-full.txt