>>

Run ERNIE-4.5-VL-28B-A3B-Paddle on Lium

Baidu · Vision and text · 29.4B parameters · 58.8 GB 8-bit weights · released 2025-06-28
It needs about 71 GB of VRAM. The cheapest setup that runs it is 4x RTX 3090 (96 GB), from $0.60 per hour, billed per second.

Other GPUs that run it

A100 80GB PCIe80 GBfrom $0.81 / hourRent
2x RTX A600096 GBfrom $0.84 / hourRent

How to run it

Rent the pod above, open a terminal over SSH or Jupyter and serve baidu/ERNIE-4.5-VL-28B-A3B-Paddle:
vllm serve baidu/ERNIE-4.5-VL-28B-A3B-Paddle --tensor-parallel-size 4
Claude Code and Codex can do all of it through the Lium CLI.

Frequently asked questions

How much VRAM does ERNIE-4.5-VL-28B-A3B-Paddle need?

About 58.8 GB for the released 8-bit weights, so plan on 71 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 ERNIE-4.5-VL-28B-A3B-Paddle?

4x RTX 3090 (96 GB). On Lium that setup costs from $0.60 per hour, billed per second with no minimum.
All open models
For AI agents (CLI install, agent skill, headless signup): llms-full.txt