>>

Run Kimi-VL-A3B-Instruct on Lium

Moonshot AI · Vision and text · 16.4B parameters · 32.8 GB BF16 weights · released 2025-04-09
It needs about 40 GB of VRAM. The cheapest setup that runs it is 2x RTX 3090 (48 GB), from $0.40 per hour, billed per second.

Other GPUs that run it

RTX A600048 GBfrom $0.42 / hourRent
4x Tesla V10064 GBfrom $0.84 / hourRent

How to run it

Rent the pod above, open a terminal over SSH or Jupyter and serve moonshotai/Kimi-VL-A3B-Instruct:
vllm serve moonshotai/Kimi-VL-A3B-Instruct --tensor-parallel-size 2
Claude Code and Codex can do all of it through the Lium CLI.

Frequently asked questions

How much VRAM does Kimi-VL-A3B-Instruct need?

About 32.8 GB for the released BF16 weights, so plan on 40 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 Kimi-VL-A3B-Instruct?

2x RTX 3090 (48 GB). On Lium that setup costs from $0.40 per hour, billed per second with no minimum.
All open models
For AI agents (CLI install, agent skill, headless signup): llms-full.txt