How to run PixArt-Sigma XL/2 1024-MS locally

PixArt-alpha (Huawei Noah's Ark Lab) · Text → image · 600M params · OpenRAIL++ (commercial OK)

An efficient 0.6B weak-to-strong DiT that reaches near-SDXL quality at 1024px (and up to 4K) at a fraction of the parameters and compute.

What it costs to run — $0 markup

On your own machine — $0. Runs free if you have about 8.0 GB of VRAM (~8GB; very small transformer, can offload further), via Diffusers / ComfyUI.
Rent a GPU — from $0.26/hr. Fits RTX 3090 24GB at the direct vendor price ($0 markup) — pay only for the minutes you generate.
Download the weights — free. Open weights at PixArt-alpha/PixArt-Sigma-XL-2-1024-MS.

Note: generative-media models are billed per image / per second / per minute on hosted services — not per token. Running locally or on your own rented GPU is usually far cheaper and keeps your data on your machine.

Key facts

Does	Text → image
VRAM to run	~8.0 GB (~8GB; very small transformer, can offload further)
Download	~2.4 GB
Parameters	600M
License	OpenRAIL++ (commercial use OK)
Run with	Diffusers / ComfyUI

Get PixArt-Sigma XL/2 1024-MS on Hugging Face →

More image models

FLUX.1 [dev] — Text → image, ~12 GB VRAM
Stable Diffusion XL 1.0 (SDXL) — Text → image, ~8.0 GB VRAM
Stable Diffusion 1.5 — Text → image, ~4.0 GB VRAM
FLUX.1 Kontext [dev] — Image editing, ~12 GB VRAM
FLUX.1 [schnell] — Text → image, ~12 GB VRAM
Qwen-Image — Text → image, ~24 GB VRAM

Browse: all media models · chat / LLM models